If you own a gaming rig, there’s without a doubt a big chunky fan in the middle of your computer just sitting there blowing air. That’s attached to what's called a CPU air cooler, and while even the most rookie of PC enthusiasts know that its job is to cool down your CPU and keep temperatures in check, the majority of us have never really given much thought to how that massive block of cooling hardware actually comes together behind the scenes.

Well, it seems like today’s the day. Over on the YouTube channel SatisFactory Process, which is all about China’s top manufacturing facilities, a video has been uploaded showcasing DeepCool’s main production facility in Beijing. It gives complete insight into how some of their premium CPU air coolers are put together, and shows how raw materials, precision machinery, and the human touch all come together to help bring your CPU temps down.

The manufacturing process starts with copper tubes, which form the backbone of today’s coolers, whether for a CPU or a graphics card. Often called 'heatpipes,' these are made from scratch entirely on site, and it’s the part of the process that involves the most significant use of machinery. And that’s just step one. There are over 15 different steps, including automated filling, precision cutting, multi-stage cleaning, and packaging, each one as intricate and complex as the next.

Manufacturing a top-end CPU air cooler takes more than just machines

PC components maker DeepCool has taken the wraps off some of the treats it has in store for Computex 2025, which kicks off next week. It headlines a pre-show announcement with its Assassin VC Elite WH vapor chamber-enhanced CPU air cooler. However, perhaps more interesting and on-trend is the confirmation that it has added wood panel options to its AK G2 CPU air coolers. There’s a couple of new AiO liquid coolers and PC cases too.

Respecting Deepcool’s announcement order, let’s look at the Assassin VC Elite WH vapor chamber-enhanced CPU air cooler first. Heralded as an air-cooler “for hardcore gamers and premium builds,” DeepCool says that this new design is a good choice for high-TDP CPUs from both AMD and Intel. It will handle them with “whisper-quiet performance” too, it asserts.

The WH suffix on this model name indicates this is a white model which, for some reason, DeepCool seems to have announced ahead of any black option. DeepCool’s first vapor chamber (VC) air cooler design was a black model, though - the Assassin IV VC Vision (with digital display). We reviewed the DeepCool Assassin IV (not a vapor chamber design) in July 2023.

If you are interested in the new Assassin VC Elite WH, there’s already a fully populated product page on DeepCool’s site (linked above). We don’t have pricing or availability nailed down yet.

More timber touches for PC builders

DeepCool seems pretty proud of its AK series CPU air coolers, which have become well established in the PC DIY community, even though they debuted as recently as 2021.

Now it is ready to announce the second-generation AK G2 air coolers. These some with “upgraded thermal control alongside 0-RPM Startup with Activate Clearing Tech (ACT), AI Smart Mode for noise-performance balance, a refined design, and an offset mounting solution supporting Intel LGA1851 and AM5,” explains the firm.

Strangely, though DeepCool has been teasing the wood finish AK designs on its worldwide social media accounts, its PR doesn’t even mention the new finish option. It is there to see, though, and in the gallery above you will also see some AK G2 digital display models that are on the way. Seems like if you want wood, you can’t also have the digital display, which is a shame.

We don’t see DeepCool AK G2 product pages yet, so we will have to be patient for a fuller reveal, availability, and pricing info - perhaps next week.

Sticking to CPU cooling products for now, DeepCool is introducing two new flagship AiO liquid coolers. The new Spartacus model comes with what DeepCool says is a “massive” LCD, at 3.4-inches (sorry, that isn’t that big DC). But this premium aluminum constructed design comes in either 360mm or 420mm models, which is admittedly sizable.

Its other new AiO is the LQ Ultra, equipped with an ARGB Halo pump, which also includes an LCD status / monitoring display. This one is going to be available in 240mm and 360mm models.

Liquid cooling cases

DeepCool also has a couple of new PC cases lined up for Computex 2025. These are purposely designed to accommodate liquid cooling apparatus, namely the bulky radiators, which are standard in such designs. If you wanted to fit something like the aforementioned Spartacus AI at 420mm, you would have to opt for the new Genome III.

The Genome III is the successor to the II, which came out in 2016. Key features include its liquid cooling reservoir, which runs the height of the case, visible in the front panel. The design also features a 5.5-inch display built into the PSU shroud and easily glanceable through the glass side panel.

After we wrote last week about DeepCool's sanctioned cooling products being sold on Amazon under the Shaking Tank name (with lightly blurred DeepCool logos), it became clear that the Amazon seller was also hawking products from competing companies, most notably Cooler Master, as pointed out by YouTuber Greg Salazar, who broke the story originally.

With that revelation, it was already pretty clear that DeepCool (sanctioned in June for selling products to two Russian companies supporting the war against Ukraine) was not directly involved in the rebranding of its products, or trying to get around U.S. sanctions to sell a few of its coolers here.

But as the company is already dealing with major fallout over the above sanctions, a representative reached out to us to clarify that it was not involved in the Saking Tank shenanigans on Amazon, and it has since taken direct action with the seller to make sure its products are no longer offered.

"It was not a DeepCool official action at all," a company representative wrote via email. "We have taken immediate action to investigate through our distribution channel and found it was one of our Chinese distributors who has been selling DeepCool products with blurred logo and brand name changed."

The DeepCool representative states that the distributor violated their contract, is being penalized, and has been ordered to remove the company's products from its overseas sales platform. As of this writing, it appears that there are no longer any renamed DeepCool products available on the seller's page. The blurred-out Cooler Master products seem to have been removed as well, though there are still several clearly labeled products from Cooler Master, Coolleo, and Juishark available via the Shaking Tank page on Amazon.

As DeepCool's representative points out, the presence of the above products on the Shaking Tank seller page "also proves… that there's no way we renamed ourselves to merchandise competitors' products."

This makes sense both from a logical and a business standpoint. It's unlikely that the small amount of profit the company could make from selling a few coolers from a third-party Amazon seller would be worth further fallout in the U.S.-aligned countries where its products are still available, or jeopardizing its presumed long-term hopes to re-enter the U.S. market. 

On Wednesday, the U.S. State Department announced sanctions on Beijing DeepCool Industries, a popular manufacturer of PC cases, cooling devices, and power supplies. DeepCool is alleged to have supplied Russia with more than $1 million worth of goods on the Common High Priority List (CHPL), which could help it in its war efforts against Ukraine.

According to the press release, DeepCool sold goods to two Russian companies known to support war efforts. These companies are Aktsionernoe Obshchestvo Taskom and OOO Novyi Ai Ti Project. Taskom is involved in freight transport, and OOO Novyi is a PC manufacturer.

The Bureau of Industry and Security, a division of the U.S. Department of Commerce, maintains the CHPL. This is a collection of items it says “pose a heightened risk of being diverted illegally to Russia because of their importance to Russia’s war efforts.”

DeepCool recently showcased new products at Computex 2024, including air and liquid CPU coolers, fans, PC cases, and power supplies compatible with the latest PC hardware. However, the company’s addition to the list of sanctioned companies means it won’t be able to sell those products in the U.S.

DeepCool’s U.S. subsidiary may have to stop all sales and most other operations immediately. This could affect its ability to provide after-sales support to customers. The sanctions mean it cannot transact any business with its parent company in China. By law, U.S. residents can no longer conduct business with the company, either. 

At this time, retailers like Amazon, MicroCenter, and Newegg are still listing DeepCool products for sale. How long that remains true is up in the air. According to the press release, “All transactions by U.S. persons or within (or transiting) the United States that involve any property or interests in property of designated or otherwise blocked persons are prohibited unless authorized by a general or specific license issued by OFAC or exempt.”

In other words, if I buy a DeepCool power supply today, I’m technically breaking the law. I might be able to make the case that I’m conducting business with a U.S. company, not the listed company, but such a claim would likely be tenuous at best.

DeepCool showed a bunch of new hardware at Computex 2024, with new air and liquid coolers for CPUs, plus fans, PC cases, and power supplies that comply with the latest Intel ATX 3.1 and PCIe 5.1 standards. DeepCool has options ready for the newly introduced standards, including motherboards with reversed side connectors. Additionally, it expanded on a couple of its CPU cooler models while also providing some decorative options that will appeal to aesthetically minded users.

CPU coolers for air and water 

DeepCool says it improved its coolers compared to the previous iteration and introduced the Assassin IV VC Vision CPU cooler that has a vapor chamber in its base. This provides for better heat transfer and it's capable of cooling CPUs rated with for up to 300 watts TDP. The cooler also has a built-in display to show data about CPU usage, temperature, power draw, and clock speed.

DeepCool has had its AK620 and AK400 CPU coolers with similar display systems in the market available for a while, and we already reviewed the former. The difference between these is that the Assassin IV VC has a vapor chamber on the heat base. A few months ago, we reviewed the DeepCool Assassin 4S CPU cooler that doesn't have the display or the vapor chamber, so it will be interesting to see how much of an improvement it provides.

Additionally, DeepCool has new low-profile AN400 series CPU coolers to ensure adequate cooling for compact PC builds, with a choice of nickel-plated and black-anodized heatsinks. There are no bells or whistles, which means no RGB something that many HTPC builders will appreciate.

In related news, Noctua introduced its NH-L12Sx77 low-profile cooler recently that allows clearance for tall RAM sticks and heatsinks. These low-profile CPU coolers are useful for mini-ITX system builders as they can provide good enough cooling within the space constraints, as many larger coolers won't fit within diminutive ITX cases.

For liquid cooling, DeepCool introduced its LP Matrix AIO series that has a dot matrix screen on its waterblock, along with a fifth generation pump and ARGB effects on the fans and the sides of the display. It's available in both 240mm and 360mm form factors.

The dot-matrix screen on these blocks seems to be customizable, more than what you can do with typical air cooler displays. Many people may not find it useful to display the CPU status all the time. Being able to show other graphics may appeal to such users, letting them make their PCs stand out.

PC chassis, accessories and new power supplies

For BTF form factor motherboards coming from Asus, MSI, and Gigabyte, DeepCool showcased its CH560R series PC cases that provide the needed compatibility with the back-side connectors. Elsewhere, the CH160 Mesh ITX case series provides space for taller air coolers and some of the larger graphics cards. To complement its mini-ITX footprint, it also has monitoring screens. Both cases should provide ample airflow for all the components, though noise could be a concern.

DeepCool also added its digital monitoring screen to its mini-ITX case, creating the CH170 Digital ITX that can be seen toward the bottom of the front panel. The CH160 Mesh ITX case also shows the 'Pixels' bits that allow you to install silicone molds made by DeepCool the company. These can be installed in currently available cases and will also be compatible with future models.

DeepCool also showed newer iterations of its 120mm and 140mm fans called the FD12 and FD14 series, available in black and white, as well as with or without ARGB support. Like many other fan makers, DeepCool uses its own proprietary 8-pin daisy chain cables that also allow lighting control.

The PN-M series power supplies now has a 12V-2x6 connector for Nvidia graphics cards, along with support for up to five EPS and PCIe connectors. The PX-S is an SFX power supply that also includes a 12V-2x6 cable with variants up to 1,000 watts. Both series support Intel ATX 3.1 and PCIe 5.1 standards.

DeepCool cases and power supplies are designed to be future-ready, with support for the latest motherboards designs and power supply standards. The company has been making PC accessories for a while now and tends to do a good job, and the new ability to customize cases with silicone bits may appeal to a certain demographic. We look forward to seeing the retail products over the coming months.

DeepCool has unveiled a new portable Mini-ITX chassis with more mesh paneling than a typical ATX chassis. Dubbed the CH160, the Mini-ITX case sports mesh paneling on the top, front, rear, bottom, and right side panel to maximize airflow. The only part of the case that is not mesh is the left side panel, which is tempered glass.

The CH160 follows a new trend in high-performance airflow-optimized cases where right-side mounted mesh panels are being installed in place of solid side panels, to boost airflow potential as much as possible.

The case measures 13.23 x 7.87 x 11.16 inches and supports up to 172mm tall air towers and 305mm triple-slot graphics cards. The power supply is mounted in the front of the chassis and supports both ATX and SFX PSU. However, with an SFX unit, the case is far more roomy, with enough room to make a front intake fan viable. An ATX unit will also impact CPU temperatures if a full-sized 172mm air cooler is installed. If you do go with an ATX unit, it would probably be best to use a low-profile air cooler or a 120mm AIO.

The only case fan size the CH160 supports is 120mm. A single fan can be added to the front (with an SFX PSU), two at the top and one in the rear. The front 120mm can be swapped out for a 2.5-inch SSD/ 3.5-inch HDD drive mount for additional storage. And this storage mount is the only form of off-board storage the case supports. So you'll need to use M.2 storage on the motherboard itself for any other storage needs.

Front I/O features two USB 3.0 ports, a Gen 2 Type C port, and a single audio jack. Deep Cool has also included a carrying handle on the top, making it easy to transport the case around. The CH160 comes in black and white color schemes.

Pricing has not been announced, but we expect this to be a cheaper case, since it does not include any case fans.

Investing in a quality cooler is a smart move, and you definitely need the best CPU cooler for your money. If you're shopping on a budget, however, it can be a little harder to find something substantial for the right price point. That's when you need deals like this one on the DeepCool AS500 Plus CPU air cooler. It usually goes for around $69 but right now at Amazon, you can take it home for just $44.

So far, no expiration has been specified for the offer so it's not clear for how long it will be made available at this price. If you're curious how it stacks up, you may want to check out our review of the DeepCool AS500 cooler, it's a little different from the Plus edition but we were very pleased with our experience.

The DeepCool AS500 Plus is a 12V, 1.32-watt cooler. It uses a TF140S FDB fan which can reach up to 1200 RPM and can get as loud as 29.2 dB. The fan also has PWM support to adjust the fan speed as the CPU temperature changes.

This cooler has a 48mm heatsink, supported by six separate heat pipes to help disperse heat. There is also a top cover with RGB LEDs should you decide to spruce up your rig with some flash and style.

If you want to get a closer look at this offer for yourself, visit the DeepCool AS500 Plus CPU air cooler product page at Amazon for more details and purchase options.

DeepCool is well known in the cooling and component world, having delivered innovative, interesting and budget-conscious PC products for more than a decade. Its current lineup includes air and AIO coolers, computer cases, keyboards, power supplies and other accessories.

Last summer we tested DeepCool’s Assassin IV, a cube-shaped cooler with a unique design, and found it to be one of the best CPU coolers on the market. Today, we’re looking at a reduced-price follow-up to the Assassin IV, the Assassin 4S. Does it have what it takes to stand out among excellent budget options like the Phantom Spirit 120 EVO? We’ll have to look at our benchmark results to say for sure. First, let’s look at the specifications of the cooler and what makes it different from the original version of the Assassin IV.

Cooler specifications

Packing and included contents 

The heatsink is protected during shipping by molded cardboard and the accessories arrive in a compact brown box.

Included in the box are the following:

• Heatsink
• Mounting for modern AMD and Intel Platforms
• 1x 140mm fan
• DeepCool DM9 Thermal Paste
• Paste spreading tool
• Mounting accessories
• Instruction Manual
• Screwdriver

LGA 1700 installation

The installation of the cooler is unchanged from the original version of this cooler, and it’s simple to set up.

1. Press the backplate against the rear of the motherboard and then secure it using the included metal standoffs.

2. Place the mounting bars on top of the standoffs and secure them with the included thumbscrews.

3. Next, you’ll need to apply thermal paste to the CPU. If you're unsure how to do that, see our How to Apply Thermal Paste primer.

4. Remove the fan from the heatsink and place the cooler on top of the CPU.  Secure it using the screws in between the two towers, using the included screwdriver.

5. Secure the fan on the cooler using the included fan clips, and then connect the fan to the PWM header on your motherboard.

Features of DeepCool’s Assassin 4S

▶ Six-year warranty

Many manufacturers include just a two- or three-year warranty with their air coolers, a potential red flag that the manufacturer doesn’t have faith in the quality of its design. DeepCool, however, is fully confident in the quality of the Assassin 4S – backing it with an impressive six-year warranty.

▶ Matrix Checkerboard Fin Design

Like its predecessor and the AK series of air coolers, the Assassin 4S incorporates a checkerboard matrix fin design that’s said to help improve total static pressure. 

▶ Complete RAM compatibility

DeepCool’s Assassin 4S doesn’t overhang DIMM slots in any manner, allowing for complete compatibility with all DDR4 and DDR5, no matter how tall your sticks are.

▶ Upgraded 140mm fan

There’s more to a cooler than just the heatsink and radiator. The fan or fans included have a large impact on a cooler’s performance and are directly responsible for how loudly it runs. In the case of the Assassin 4S, the fan is pretty much the only thing different in comparison to the previous model. There is only one 140mm spinner included with the Assassin 4S, and it has been upgraded for improved performance. It has seven larger fan blades instead of the nine slimmer fan blades included with the original model, and features improved static pressure ratings. 

One complaint I have about the Assassin 4S is that DeepCool does not include the back fan bracket included with the Assassin IV, and it is not available to purchase separately – which means you won’t be able to add an additional fan, unless you’re willing to 3D print a custom mount. 

LGA1700 Socket Bending

There are many factors other than the CPU cooler that can influence your cooling performance, including the case you use and the fans installed in it. A system's motherboard can also influence this, especially if it suffers from bending, which results in poor cooler contact with the CPU. 

To prevent bending from impacting our cooling results, we’ve installed Thermalright’s LGA 1700 contact frame into our testing rig. If your motherboard is affected by bending, your thermal results will be worse than those shown below. Not all motherboards are affected equally by this issue. I tested Raptor Lake CPUs in two motherboards. And while one of them showed significant thermal improvements after installing Thermalright’s LGA1700 contact frame, the other motherboard showed no difference in temperatures whatsoever! Check out our review of the contact frame for more information.

Testing methodology

Today's highest-end CPUs, whether Intel or AMD, are difficult to cool in intensive workloads. In the past. reaching 95 degrees Celsius or more on a desktop CPU might have been a cause for concern. But with today’s top-end CPUs, this is considered normal operation. Similar behavior has been present in laptops for years due to cooling limitations in tight spaces.

All testing is performed with a 23C ambient room temperature. Multiple thermal tests are run on each CPU to test the cooler in a variety of conditions, and acoustic measurements are taken with each result. These tests include:

1. Noise normalized testing at low noise levels

2. “Out of the box”/default configuration thermal and acoustics testing

     a. No power limits enforced

     b. Because CPUs hit Tjmax in this scenario, the best way to compare cooling strength is by recording the total CPU package power consumption.

3. Thermal and acoustics testing in power-limited scenarios

     a. Power limited to 175W to emulate a medium-intensity workload

     b. Power limited to 125W to emulate a low-intensity workload

The thermal results included are for 10-minute testing runs. To be sure that was sufficiently long to tax the cooler, we tested both Thermalright’s Assassin X 120 R SE and DeepCool’s LT720 with a 30-minute Cinebench test with Intel’s i9-13900K for both 10 minutes and 30 minutes. The results didn’t change much at all with the longer test: The average clock speeds maintained dropped by 29 MHz on DeepCool’s LT720 and 31 MHz on Thermalright’s Assassin X 120 R SE. That’s an incredibly small 0.6% difference in clock speeds maintained, a margin of error difference that tells us that the 10-minute tests are indeed long enough to properly test the coolers. 

Testing configuration – Intel LGA1700 platform 

Thermal results without power limits

Without power limits enforced on Intel’s i7-13700K, the CPU will hit its peak temperature and thermally throttle with even the strongest of air coolers. When the CPU reaches its peak temperature, I measure the CPU package power to determine the maximum wattage cooled to best compare their performance. The results below do not include the best liquid coolers on the market, which are able to keep the CPU under TJMax (100 degrees Celsius).

The Assassin 4S is nearly as good as the original Assassin IV, trailing it by 3-4W in cooling performance – but it runs more quietly. In its full-speed mode, its sound output is more than 3dBA lower! The low-noise mode doesn’t show a drastic change from the previous cooler, only improving by 0.7 dBA over the low-noise mode tested on the Assassin IV.

Thermal results with noise-normalized to 38.2 dBA

Finding the right balance between fan noise and cooling performance is important. While running fans at full speed can improve cooling capacity to some extent, the benefits are limited and many users prefer a quiet system. With this noise-normalized test, I’ve set noise levels to 38.2 dba. This noise level is a low volume level, but slightly audible to most people.

The Assassin 4S narrowly outperforms its predecessor by two watts when noise normalized, showing the improvements from the upgraded fan design.

175W Cinebench results

Most coolers on the market can keep Intel’s i7-13700K under its peak temperature if the power consumption is limited, so for this test we’ll be looking at the CPU’s actual temperature.

With a steady 175W load run on the CPU, the temperature averaged 56 degrees C over the ambient room temperature of 23 C in full-performance mode and 58 C in low-noise mode. This is another extremely impressive result, tied for the second-best result from any air cooler I’ve tested thus far, and an improvement of 2 C compared to its predecessor.

The 4S’s thermal performance is equal to the original Assassin IV in low-noise mode, and only a single degree behind in full performance mode. This is especially impressive when you consider noise levels – the Assassin 4S runs at only 39.6 dBA, compared to the 43.4 dBA produced by the Assassin IV.

125W Cinebench results

The lowest power limit I test with Raptor Lake CPUs is 125W. This is a high enough limit to allow the CPU to maintain its base clock speeds even in the most intensive tests, and most coolers should be capable of keeping the CPU below TJ Max (the max temperature before throttling) – even low-end coolers.

Thermals don’t really matter in this scenario. Even Intel’s stock cooler can handle a load like this with ease. Noise levels, rather than CPU temperature, are the important factor here. These were ideal, never overpowering the ambient system noise of 37.3 dBA in both full performance and low noise modes. What this means is that the cooler ran quieter than my system fans.

Conclusion

The Assassin 4S offers a unique, arguably attractive aesthetic unlike any other save those that came before it in the same line. And it delivers quieter noise levels than its predecessor. While it is cheaper than the Assassin IV, its price tag of $79.99 puts it firmly into “premium” territory and makes it hard to recommend for most users.

If it were just a bit cheaper, I’d have an easier time recommending it, and it will likely go on sale at some point, so keep an eye out if you're interested. DeepCool backs the Assassin 4S with an impressive 6-year warranty – twice as long as many competitors – which makes the higher price tag a little bit easier to swallow.

Cooling specialist DeepCool has released three new CPU air coolers based on its existing tried and trusted tower form factor product stack. Obviously, the new DeepCool AK400 Digital, AK500 Digital, and AK620 Digital are new versions of the AK400, AK500 and AK620. What makes these the new Digital Series is the addition of "a crystal-clear digital screen" to the top of the tower heatsinks.

LCD screens are becoming the early-2020s answer to RGB bling. However, they aren't quite as superfluous, with most component makers demonstrating that the screens are useful for real-time system monitoring information. DeepCool follows this trend, and on all of these new Digital Series coolers, you can configure the display using software to show CPU temperature (Celsius or Fahrenheit), CPU utilization, or both of these variables - in real time. DeepCool's accessory software also has a built-in alarm function.

We don't have any specs for the screens on these new coolers. However, they appear to be quite simple calculator-style LCD displays with backlighting. So you aren't going to be able to send any videos or animations that you might like to play on them.

Just because these new coolers have status monitor displays on the top doesn't mean that ARGB lighting has been cast aside. All three new models also feature ARGB LED strips that can be controlled by all the popular supported standards like AURA Sync, RGB Fusion, Mystic Light, etc.

The entry level AK400 Digital is the slimmest of the new bunch, equipped with a single FDB (fluid dynamic bearing) FK120 fan. Rated for cooling up to 220 W, this cooler is going to easily take a non-overclocked 65 W TDP processor in its stride, and probably more. Surprisingly, the larger AK500 Digital's cooling capacity is only claimed to be 20 W more, and the sizable twin-fan AK620 Digital adds another 20 W to its rating, with claims that it is capable of dissipating 260 W. For some insight into how these Digital Series coolers should perform, please check out both our DeepCool AK500 review, and our DeepCool AK620 review - both products scored very well.

As modern cooling products, these new DeepCool AK Digital Series designs are compatible with the latest Intel LGA 1700 and AMD AM5 sockets as well as older designs like LGA 1200 / 1151 / 1150 / 1155 and AM4. DeepCool says that its all-metal mounting bracket is sturdy, and easy to use with a simple five step fixing process.

There are some drawbacks to these Digital Series CPU air coolers to make you aware of. Firstly, we note that these designs command an approximate $10 price premium. We have also discussed some limits to the capabilities of the built-in displays, above. Lastly, we must highlight the requirement for three cables attached to the CPU cooler to access the full gamut of functions (fan, ARGB lighting, and display).

The AK Digital Series was originally revealed in China in mid-July. When we checked today we found they are listed at retail in the US, with Amazon pricing of the DeepCool AK400 Zero Dark Digital Display Edition at $52.60, ranging up to the DeepCool AK620 Zero Dark Digital Display Edition at $79.98. These prices don't look great compared to the Euro prices previously announced, so it might be worth waiting for more buying options. If you would like a white version of one of these AK Digital Series coolers, they are expected to become available in September.

DeepCool may not be the most well-known name in the cooling and component world, but the company has been delivering innovative, interesting and budget-conscious PC cases and cooling products for more than a decade. Its current lineup includes air and AIO coolers, computer cases, keyboards, power supplies and other accessories. We’ve been impressed recently by the company’s AK620 and Assassin III air coolers, and DeepCool’s 360mm LT720 AIO currently holds our top recommendation on our list of best AIO coolers.  

The company’s latest cooler to land on our test bench is the DeepCool Assassin IV. The company says it has been years in the making, with the aim of achieving a new pinnacle in air cooling aesthetic and performance. We’ll of course have to put it through our usual testing to see if it achieves those lofty goals to earn a spot on our best CPU coolers list, but here are the cooler’s full specs, direct from DeepCool.

Specs

Packing and Included Contents 

DeepCool’s Assassin IV arrives in a box that is slightly smaller than, but similar in size to, other coolers in its class.

The inner contents are protected by molded foam and cardboard, insuring the cooler arrives undamaged.

Included with the package are the following:

• Screwdriver
• Manual

Installation on LGA1700 and AMD AM4 AM5

The setup and installation of DeepCool’s Assassin IV is fairly simple. The process is similar for both AMD Ryzen and Intel motherboards: 

1. If you’re running an AMD Ryzen system, you’ll need to start by removing the default retention bracket. Intel users will need to apply the backplate to the motherboard.

2. Secure the mounting standoffs on the mounting holes.

3. Place the mounting bars on top of the standoffs and secure them with the included thumbscrews.

4. Remove the middle top cover of the Assassin IV and remove the 140mm fan underneath.

5. Place the Assassin IV on top of the mounting bars and use the included screwdriver to secure it to the bars.

6. Replace the middle fan and top cover, then connect the PWM cables to the motherboard.

Features of DeepCool’s Assassin IV Air Cooler

The latest Assassin features a new eye-catching design. It not only looks good, but was built for not just top-tier cooling performance, but the best performance while running quietly. Does it have what it takes to compete with coolers like Noctua’s NH-D15 and Cooler Master’s MA824 Stealth? We’ll look at performance in the benchmarks section, but first, let’s talk about some of the features of the Assassin IV that make it stand out from other coolers.

Change from quiet and performance modes at the flick of a switch

The Assassin IV supports two modes of operation out of the box: quiet and performance modes. The quiet mode, as its name suggests, implements a fan curve tuned for silence and lowers the maximum fan speed. Extreme mode allows for unrestricted operation for maximum cooling capacity, at the cost of slightly higher noise levels.

Checkerboard Matrix fin design

The Assassin IV incorporates DeepCool’s unique checkerboard matrix design, which aims to maintain increased levels of air pressure.

Full RAM compatibility

In the default configuration, DeepCool’s Assassin IV provides full compatibility with any size of RAM module, as the cooler does not overhang DIMM slots at all, at least in its standard configuration. More on that in the very next section.

Solid block design, mount for an optional third fan

The Assassin IV features a “solid block” style design, reminiscent of a Borg Cube from Star Trek. This design improves air pressure by eliminating potential pathways for air leakage. There’s no place for fan clips in this design, as the fans are connected via frames instead.

The rear frame can be moved to the front, or you can add an additional fan (not included) using an extra fan frame, included with the unit. However, if you use three fans you’ll need to make sure your RAM is less than 42mm tall.

Two newly designed FDB fans, one 120mm and one 140mm

There’s more to a cooler than just the heatsink or radiator. The bundled fans have a significant impact on cooling and noise levels. The Assassin IV comes with two new FDB fans of different sizes pre-installed: a 140mm fan in the middle and a 120mm fan at the rear.

These fans incorporate fluid dynamic bearings and a newly designed 3 phase, six-stator motors with a focus on reduced power consumption.

Seven Staggered Heatpipes, Precision-Machined Convex Copper Base

The Assassin IV includes seven 6mm heatpipes spread across two heatsinks. The first heatsink is smaller, with the heatpipes arranged in a uniform order. The second heatsink is larger, with heatpipes staggered for optimal heat dissipation.

DM9 Industrial Grade Thermal Paste

The heatsink and the fans aren’t the only things that impact cooling. The thermal paste included also has an impact on a cooler’s operation. DeepCool includes its DM9 Industrial Level thermal paste with the Assassin IV, along with cleaning wipes and a thermal paste spread tool. The paste is said to be composed of a highly stable silicon resin compound which is not electrically conductive nor capacitive – avoiding the problems posed with liquid metal.

Testing

Modern CPUs, whether Intel or AMD, are difficult to cool in intensive workloads. In the past reaching 95C+ on a desktop CPU might have been a cause for concern – but with today’s high-end CPUs, this is considered normal operation. Similar behavior has been present in laptops for years due to cooling limitations in tight spaces. 

Since last fall, Tom’s Hardware has brought you cooling reviews using one of the most power-hungry desktop CPUs on the market – Intel’s flagship i9-13900K. To give you an idea of what it takes to cool Intel’s behemoth, we’ve tested it with a variety of coolers from entry level low-end air coolers like the Amazon Basics CPU cooler, to high-end 420mm AIOs such as Corsair’s iCUE H170i Elite.

While it’s nice to see how Intel’s flagship responds to different levels of cooling, those results don’t always correlate with lower-tier CPUs. Today’s review features two lower-end (though still powerful) CPUs  – AMD’s Ryzen 7 7700X and Intel’s i7-13700K.

Testing Methodology

All testing is performed with a 23C ambient room temperature. Multiple thermal tests are run on each CPU to test the cooler in a variety of conditions, and acoustic measurements are taken with each result. These tests include:

1. Noise-normalized testing at low noise levels

2. Out-of-the-box/Default Configuration thermal & acoustics testing

     a. This means no power limits on Intel’s i7-13700K, and AMD’s default power limits on AMD’s Ryzen 7 7700X.

     b. Because CPUs hit Tjmax in this scenario, the best way to compare cooling strength is by recording the total CPU package power consumption.

3. Thermal & Acoustics testing in Power Limited Scenarios

    a. With Ryzen 7 7700X, I’ve tested with limits of 95W and 75W enforced.

    b. On Intel’s i7-13700K, I’ve tested with limits of 175W and 125W enforced.

The thermal results included are for a 10-minute testing runs. To be sure that was sufficiently long to tax the cooler, we tested both Thermalright’s Assassin X 120 R SE and DeepCool’s LT720 with a 30-minute Cinebench test with Intel’s i9-13900K for both 10 minutes and 30 minutes. The results didn’t change much at all with the longer test: The average clock speeds maintained dropped by 29 MHz on DeepCool’s LT720 and 31 MHz on Thermalright’s Assassin X 120 R SE. That’s an incredibly small 0.6% difference in clock speeds maintained, a margin of error difference that tells us that the 10-minute tests are indeed long enough to properly test the coolers.

Testing Configuration – Intel LGA1700 Platform

Testing Configuration – AMD AM5 Platform

Intel Core i7-13700K thermal results with noise normalized to 38.2 dBA

Finding the right balance between fan noise levels and cooling performance is important. While running fans at full speed can improve cooling capacity to some extent, the benefits are limited and many users prefer a quiet system.  

In this test where the fans have been set to low noise levels, DeepCool’s Assassin IV ties for Cooler Master’s new Stealth for the best result we’ve seen on any air cooler – cooling 222W when paired with Intel’s i7-13700K!

No Power Limits Thermal Results

Without power limits enforced on Intel’s i7-13700K, the CPU will hit its peak temperature and thermally throttle with even the strongest of air coolers. In this test, we measure the total amount of watts the cooler is able to dissipate from the CPU.

Here we have another chart-topping result, with DeepCool narrowly edging past both Noctua’s NH-D15S and Cooler Master’s MA824 Stealth with 236W cooled during the course of testing. Switching to the low noise mode reduces total cooling capacity slightly, with an average of 227W cooled. This outperforms both the Peerless Assassin and Montech’s D24 Premium. 

The maximum noise levels of the Assassin IV are also very good. It’s slightly quieter than Noctua’s NH-D15S and barely louder than Cooler Master’s MA824 stealth. At 42.9 dBA, the noise levels are audible, but far from loud.

175W Cinebench Results

Most coolers on the market are able to keep Intel’s i7-13700K under its peak temperature if the power consumption is limited, so for this test we’ll be looking at the CPU’s actual temperature.

With this slightly restricted workload, DeepCool’s Assassin IV performs very well – but runs a little hotter than Noctua’s NH-D15S. There’s not much of a difference between the full-speed and low-noise modes. But the low-noise mode runs significantly quieter. In the full-speed profile, the cooler hit its maximum noise level of 42.9 dBA in this test. Switching to the low noise mode only increases the CPU’s temperature by a single degree, but reduces the noise profile considerably – at only 38.2 dBA, it runs very quietly and is barely audible.

125W Cinebench Results

The lowest power limit I test with Raptor Lake CPUs is 125W. This is a high enough limit to allow the CPU to maintain its base clock speeds even in the most intensive tests. And even low-end coolers should be able to keep the CPU below Tjmax, the highest temperature the chips is designed to run at before throttling.

The Assassin IV again delivers chart-topping thermal performance here, and it does so while running fully silently. With this system using BeQuiet’s Silent Base 802 PC case, the minimum noise levels reach 37.3 dBA. That’s also the acoustic measurement when testing at this lower power limit, both in full speed and low noise modes, which means that the cooler was running quieter than the sound of my system’s fans. This is an excellent result for noise levels.

AMD Ryzen 7 7700X Thermal Results

It’s great to see how coolers perform with hotter CPUs like Intel’s i7-13700K, but that performance doesn’t always directly translate to how other CPUs will operate with the same coolers. Some coolers might perform better – or worse – depending on the CPU it’s paired with. So I’ve included AMD’s Ryzen 7 7700X results to show how the cooler reacts when installed in a current-gen AMD system.

Thermal Results with noise normalized to 36.4 dBA

With noise levels set to 36.4dba, I measured 126W cooled on average during the course of testing. This is the best result I’ve seen from an air cooler and is only a few watts short of the cooler’s performance when noise levels are not restricted.

Maximum Cooling Capacity with maximum fan speeds

Next, we test the cooler’s maximum cooling capacity by allowing the fans to run at full speed. In this test, DeepCool’s Assassin IV cooled 130W on average, placing it among the best results we’ve seen on an air cooler. This is only a few watts short of AMD’s default power limit for Ryzen 7700X, which allows the CPU to use up to roughly 140W.

With maximum noise levels of 42.9 dBA, the Assassin IV runs slightly louder than Cooler Master’s MA824 Stealth and slightly quieter than Noctua’s NH-D15S.

Thermals and Acoustics with a 95W Power Limit

Not every task will force the CPU to use its full potential or maximum power draw, so it’s good to test coolers in a variety of power limits. In these situations, it’s generally more important to consider the noise levels, and the Assassin IV does great in that respect with noise levels of only 39.1 dBA. It maintains this low noise level while keeping the CPU cooler than any of the other coolers we tested for this review, maintaining a lead of 4C over its nearest competitor. This is an excellent result, showing the strength of DeepCool’s design.

Thermals and Acoustics with a 75W Power Limit

With power draw reduced to 75W on AMD’s Ryzen 7 7700X, this thermal load shouldn’t be difficult for most coolers to cool. This is roughly the amount of power Ryzen 7 7700X CPU will use during gaming, and it’s also the maximum power consumption of AMD’s non-X Ryzen 7 CPU.

At 36.4 dBA, the noise levels for the Assassin IV are extremely quiet, almost silent, matching Noctua’s NH-D15S.

At this lower level of power consumption, cooling difficulty isn’t hard at all and even the most basic coolers should have no problems here. In this situation, the temperature of the CPU isn’t very important – even the most basic coolers should be capable of handling the heat of a 75W power limit.

Conclusion

DeepCool’s Assassin IV offers a compelling package with chart-topping cooling performance, balanced noise levels and full RAM compatibility – what more could you want in an air cooler? The unique form factor is reminiscent of a Borg Cube, and the Assassin IV assimilates (and dissipates) the heat of demanding workloads with ease.

This air cooler offers top-tier cooling performance balanced with low noise levels – and the quiet mode switch will be appreciated by those who demand silent operation. Air coolers of this class don’t come cheap, though. At $99, this Assassin is firmly in the premium price range – but that’s backed up by the impressive, quiet performance we’ve shown in our testing above.

DeepCool  has, in recent years, delivered some of the best CPU coolers on the market, with options like the Assassin III and AK620. DeepCool’s current lineup includes both air and AIO coolers, as well as computer cases, keyboards, power supplies, and other accessories.   

We most recently reviewed DeepCool’s budget AG400 cooler, which performed well in our tests. Today we’re going to look at the AG400’s bigger brother, the AG620. It will be the first high-end air cooler we’ve tested with Intel’s i9-13900K. DeepCool markets the AG620 as a “a dual-tower 120mm CPU cooler that boasts impressive 260W cooling power performance that's been stripped down and optimized for terrific efficiency throughout.”

Will DeepCool’s AG620 be enough of an upgrade to significantly improve cooling performance on Intel’s i9-13900K? We’ll have to put it through testing to find out. But first, here are the AG620’s specifications:

Cooler Specifications

Packing and Included Contents 

DeepCool’s AG620 comes in a somewhat small box, with molded foam and cardboard used for protection of the inner contents.

Included with the package are the following:

• Dual-tower heatsink
• 2x 120mm fans
• Mounts for Intel & AMD platforms
• Fan Splitter
• Thermal Paste

Installation

Installing the AG620 was very easy. To begin, you press the backplate to the motherboard,then secure it with the thumbscrew standoffs. Once finished, secure the mounting bars with the included screws. Apply the included thermal paste from the included packet to your CPU, then mount the CPU block and secure it with a screwdriver. The last step, of course, is to attach the fans with the included clips.  

Features of the DeepCool AG620

Matrix Fin Array

The AG620 incorporates a matrix fin array, resembling a checkerboard, which is designed to increase the static pressure of the airflow and improve overall efficiency. 

Convex Copper Base, bi-directional heat pipe technology

The AG620 features a precision-machined convex copper base and six copper heat pipes which have been optimized for peak thermal transfer, regardless of whether the cooler is installed in a vertical or horizontal case. 

2x High Efficiency Fans

There’s more to a cooler than just the heatsink and radiator. The bundled fans have a significant impact on both cooling and noise levels. Included are two 9-bladed 120mm fans which run as low as 300 rotations per minute (RPM) for complete silence at idle loads.

Testing Methodology

While it was fairly easy with previous generations of CPUs for coolers to keep the flagship i9 processor well under TJ max (the maximum temperature a CPU can sustain without throttling) in tough workloads, this is no longer realistically possible with the Core i9-13900K without extreme cooling (or enabling power limits).

When I began testing coolers on Intel’s Core i9-12900K, I found that some products that cooled the i9-10900K well struggled in some scenarios when paired with the Alder Lake CPU – and Raptor Lake is even more difficult to cool in these situations. While in the past a CPU hitting its peak temperature was cause for concern, enthusiasts are going to have to learn to accept high temperatures as “normal” while running demanding workloads with Raptor Lake and Ryzen 7000 CPUs.

Modern AMD and Intel CPUs are designed to run fairly hot without any problems – up to 95 degrees Celsius for AMD Ryzen 7000 CPUs, and up to 100 C for Intel’s Core i9-13900K. Similar behavior has been standard in laptops for years due to cooling limitations. 

Furthermore, Intel’s i9-13900K supports Adaptive Boost Technology (ABT), which allows the CPU to dynamically boost to higher all-core frequencies based on available thermal headroom and electrical conditions. This allows multi-core loads to operate at up to 5.5 Ghz if the necessary amount of thermal dissipation is there. This feature works in a way that actively seeks high temperatures: If the chip sees that it is running underneath the 100 degrees C threshold, it will increase its performance and power consumption until it reaches the safe 100 C limit, thus sustaining higher clocks (and providing better performance) for longer periods.

The increased cooling challenges posed by Raptor Lake mean that we’ve had to change some of the ways we test coolers. Some coolers were able to pass Cinebench R23 multicore testing with Intel’s 12th Gen i9-12900K when power limits were removed (although only the strongest models were able to pass that test). 

Since the Raptor Lake 13900K actively adjusts in a way that seeks its top safe temperature, we’ll compare performance instead by looking at total benchmark scores and clock speeds maintained. 

I’ll be testing Intel’s i9-13900K CPU using Asus’ TUF Gaming Z690 Gaming Plus WIFI motherboard and Cooler Master’s HAF 700 Berserker PC case, with case fans limited to 35% speeds. The motherboard’s default fan curve is used for the CPU Cooler’s fans.

In addition to testing Cinebench without power limits enforced, we’ll also be showing results when the CPU’s power consumption is limited to a more reasonable 200W. We’ll also show results at 125W for those who prefer whisper-quiet cooling, at the cost of some performance. For both of these results, we’ll show traditional delta over ambient temperature results.

We’ll also provide noise level measurements recorded using a PSPL25 Sound Meter for all three power levels tested, to compare how much noise each cooler makes in different scenarios. We expect most coolers to run effectively silently at 125W. You’ll notice that our noise level graphs start at 36 dBA - that’s because 36dBA is the noise floor (the lowest reading we can measure) in our testing environment. Keep in mind that noise measurements are logarithmic, meaning the differences between the noise levels of the coolers will be bigger than these graphs would suggest

LGA1700 Socket Bending

Note there are many factors other than the CPU cooler that can influence your cooling performance, including the case you use and the fans installed in it. A system's motherboard can also influence this, especially if it suffers from bending, which results in poor cooler contact with the CPU. 

In order to prevent bending from impacting our cooling results, we’ve installed Thermalright’s LGA 1700 contact frame into our testing rig. If your motherboard is affected by bending, your thermal results will be worse than those shown below. Not all motherboards are affected equally by this issue. I tested Raptor Lake CPUs in two motherboards. And while one of them showed significant thermal improvements after installing Thermalright’s LGA1700 contact frame, the other motherboard showed no difference in temperatures whatsoever! Check out our review of this contact frame for more information. 

Testing Configuration

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

No Power Limits Cinebench Results

With Raptor Lake’s i9-13900K pushing speeds of 5.5 GHz or more, even the strongest of coolers hit TJ max while running Cinebench R23 and other demanding scenarios. As the 13900K is designed to aim for its top safe temperature, we’ll be comparing the overall benchmark score and the CPU’s clock speeds instead. 

The results below are for a 10-minute testing run. But to be sure that was sufficiently long to tax the cooler, we also retested both Thermalright’s Assassin X 120 R SE and DeepCool’s LT720 with a 30-minute Cinebench test. The results didn’t change much at all with the longer test. The average clock speeds maintained dropped by 29 MHz on DeepCool’s LT720 and 31 MHz on Thermalright’s Assassin X 120 R SE. Looking at this another way, that’s an incredibly small 0.6% difference in clock speeds maintained, a margin of error difference that tells us that the 10 minute tests are indeed long enough to properly test the coolers.

Because of the difficulty of cooling the i9-13900K in this workload, we feel the best way to compare coolers here is to record the average power consumption of the CPU. Looking at this by total watts cooled only, the AG620 has a cooling advantage of ~30-50W compared to lower-end air coolers. Liquid coolers can of course sustain higher amounts, but not by much. The middle of the pack AIOs cool an additional  roughly 20W, and the best performing AIO here – also made by DeepCool – can only handle up to 38W more.

Of course, total cooling capacity isn’t the only consideration in a cooler - how loud it operates is also a significant aspect. As Cinebench R23’s multi-core benchmark will push coolers to their limits, it’s also a great test for recording the worst-case scenario of fan noise levels. 

You might notice that our noise charts start at 36dB. This is the noise floor of our testing environment. This makes 36dB our baseline measure, as we’re unable to measure noise levels below this threshold. Keep in mind that noise measurements are logarithmic, meaning the differences between the noise levels of the coolers will more significant than these graphs would suggest. 

The AG620 does well here, running quieter than the other air coolers we’ve tested on the i9-13900K, with the exception of Thermalright’s SFF cooler, which has a much lower thermal capacity. It also runs quieter than all liquid coolers we’ve tested in this scenario, save for for Arctic’s Liquid Freezer II. 

I would also note that, aside from dB ratings, at full fan speeds I found the noise produced by this cooler to be slightly annoying. The pitch of the fans is whiny compared to other units, which is something that can’t be captured by my noise meter.

We’re also showing the total Cinebench R23 benchmark scores to show how little is lost  – or gained – with different cooling solutions. Between the weakest and strongest coolers, we have a total benchmark score variance of roughly 13%. 

DeepCool’s AG620 is the strongest air cooler we’ve tested thus far in this benchmark, scoring 38,849K points, which is between  between 1,300 and 3, points higher than the lower-end air coolers in this test.

200W Cinebench Results

When restricting power consumption to a more reasonable 200W, the AG620 continues its lead over the lower-end air coolers, by up to 15 degrees Celsius, but falls up to 5 degrees C behind high end 360 mm AIOs.

The acoustic performance of the AG620 was middle of the road at 200W. It wasn't particularly loud, but it wasn’t the quietest result either.

125W Cinebench Results

The lowest power limit I test at is 125W, mainly because this is also the lowest level where I can reliably measure noise measurements. Lower power consumption causes the noise of many CPU coolers to fall below the noise created by the system fans (even while restricted to 35% speed).

When limited to 125W, the AG620 ran silently; the noise profile is essentially indistinguishable from the rest of the system fans. In terms of actual CPU temperature, it’s in the middle of the pack (compared to other units), but as this is a low noise performance scenario there’s no cause for concern.

Conclusion

DeepCool’s AG620 is a solid dual tower air cooler that pairs well with Intel’s Core i9-13900K, handling loads of up to 277W with a total cooling capacity just shy of high end AIOs. At only $54.99 USD, the AG620’s value is hard to beat. But if whining fans bother you, you may want to spend more on something else, or perhaps swap out the included fans for something else with a more pleasing sound signature.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

DeepCool was founded in Beijing in 1996. And the company has, in recent years, delivered some of the best CPU coolers on the market, with options like the Assassin III and AK620. DeepCool’s current lineup includes both air and AIO coolers, as well as computer cases, keyboards, power supplies and other accessories.   

We most recently reviewed DeepCool’s budget AG400 cooler, which performed well in our tests. Today we’re going to look at the company’s newest, top-end cooling product, the 360mm LT720 AIO. It features an upgraded pump and new CPU block compared to previous generation DeepCool AIOs. But of course, the latest CPUs from AMD and Intel requiring more capable cooling, we’ll have to put the LT720 through our tests to see how well it handles the Core i9-13900K. But first, here are the LT720’s specifications from DeepCool. 

Cooler Specifications

Packing and Included Contents 

DeepCool’s LT720 comes in a fairly large box that uses plastic coverings, foam, and cardboard to protect the contents. 

Included with the package are the following:

• 360mm Radiator
• CPU block with 4th generation DeepCool water pump
• 3x 120mm fans
• 1x fan splitter
• Mounts for all modern CPU sockets (including AM5 & LGA1700)
• Information, Support, and Installation guides
• Pre-applied thermal paste

Installation

Installing DeepCool’s LT720 was simple. you press the backplate to the motherboard and then slide on the standoffs. Then you’ll need to secure the fans to the radiator and then mount the radiator to the computer case. Once that’s finished, simply press the CPU block to the socket and secure it with the included thumb screws. The last step is to attach the infinity mirror atop the CPU block.

One item of importance to note: While the LT series AIOs will function regardless of the CPU block’s orientation, the company says you should install the LT720 in the “6 O’Clock,” with it’s tubes coming from the bottom of the CPU block, for the best performance.

Features of DeepCool’s LT720

4th generation in-house CPU pump

The vast majority of Liquid Coolers on the market today are based around Asetek designs, but DeepCool’s LT720 includes a 4th generation in-house designed water pump. It features new flow channels with an optimized skived copper block, paired with a powerful 3100RPM three-phase drive pump motor for strong thermal performance.

Full Copper CPU plate 

The CPU contact plate on DeepCool’s LT720 is made of copper and comes with thermal paste pre-installed.

Infinity Mirror

On top of the water pump is an ARGB infinity mirror, which gives the LT720’s CPU block a very unique-looking piece of eye candy.

User serviceable loop 

If you ever want to manually refill the LT720, it’s possible to do so on your own. Doing so may void the warranty, but in all likelihood, the warranty is going to be expired by the time you would need to refill the loop.

Medium thickness 360mm radiator

The LT720 features a 360mm with 27mm thickness. This isn’t the thickest (or thinnest) radiator we’ve seen, but it should be compatible with most cases that have space for three 120mm fans.

3x custom DeepCool FK120 fans

There’s more to a cooler than just the heatsink or radiator. The bundled fans have a significant impact on cooling and noise levels. Included with the LT720 are 3x custom FK120 fans. The FK series is DeepCool’s line of high-performance 120mm PWM fans designed for top-tier cooling performance. But the fans included are not the same as retail FK120 fans -- these updated FK120 fans feature upgraded speeds, more airflow, and higher static pressure.

The company says high-quality polybutylene terephthalate (PBT) material is used in the production of the FK series fans to ensure increased durability and wear resistance, allowing the blades to maintain higher tensile strength. Shock-absorbing rubber pads protect the screw mounts to reduce vibrations and noise.

One feature that will be especially handy for those new to building PCs is that on each of DeepCool’s FK fans there are arrow indicators showing both the direction the fans spin and the direction of the airflow - eliminating a question that plagues many new users: Did I install the fan in the right direction? 

Testing Methodology

While it was fairly easy with previous generations of CPUs for coolers to keep the flagship i9 processor well under TJ max (the maximum temperature a CPU can sustain without throttling) in tough workloads, this is no longer realistically possible on current generation CPUs (and the 13900K especially) without extreme cooling (or enabling power limits).

While in the past a CPU hitting its peak temperature was cause for concern, enthusiasts are going to have to learn to accept high temperatures as “normal” while running demanding workloads with Raptor Lake and Ryzen 7000 CPUs. Modern AMD & Intel CPUs are designed to run fairly hot without any problems – up to 95 degrees C for AMD Ryzen 7000 CPUs, and up to 100 degrees C for Intel’s Core i9-13900K. Similar behavior has been standard in laptops for years due to cooling limitations in tight spaces. 

Furthermore, Intel’s i9-13900K supports Adaptive Boost Technology (ABT) which allows Core i9 processors to dynamically boost to higher all-core frequencies based on available thermal headroom and electrical conditions. This allows multi-core loads to operate at up to 5.5ghz if the necessary amount of thermal dissipation is there. This feature works in a way that actively seeks high temperatures: If the chip sees that it is running underneath the 100-degree C threshold, it will increase its performance and power consumption until it reaches the safe 100C limit, thus sustaining higher clocks (and providing better performance) for longer periods.

The increased cooling challenges posed by Raptor Lake mean that we’ve had to change some of the ways we test coolers. Some coolers were able to pass Cinebench R23 multicore testing with Intel’s 12th Gen i9-12900K when power limits were removed (although only the strongest models were able to pass that test). Most liquid coolers and all air coolers I’ve tested “failed” that test because the CPU reached TJ max in this scenario. 

With Raptor Lake’s 13900K, not a single cooler tested has been able to keep the CPU under TJ max in this test – because as we pointed out, the chip is designed to dial up performance and power until it richest that thermal result. We’ll compare performance instead by looking at total benchmark scores and clock speeds maintained. 

I’ll be testing Intel’s i9-13900K CPU using Asus’ TUF Gaming Z690 Gaming Plus WIFI motherboard and Cooler Master’s HAF 700 Berserker computer case, with case fans limited to 35% speeds. The motherboard’s default fan curve is used for the CPU Cooler’s fans.

In addition to testing Cinebench without power limits enforced, we’ll also be showing results when the CPU’s power consumption is limited to a more reasonable 200W. We’ll also show results at 125W for those who prefer whisper-quiet cooling, at the cost of some performance. For both of these results, we’ll show traditional delta over ambient temperature results.

We’ll provide noise level measurements recorded using a PSPL25 Sound Meter for all three power levels tested, to compare how much noise each cooler makes in different scenarios. We expect most coolers to run effectively silently at 125W.

LGA1700 Socket Bending

Note there are many factors other than the CPU cooler that can influence your cooling performance, including the case you use and the fans installed in it. A system's motherboard can also influence this, especially if it suffers from bending, which results in poor cooler contact with the CPU. 

In order to prevent bending from impacting our cooling results, we’ve installed Thermalright’s LGA 1700 contact frame into our testing rig. If your motherboard is affected by bending, your thermal results will be worse than those shown below. Not all motherboards are affected equally by this issue. I tested Raptor Lake CPUs in two motherboards. And while one of them showed significant thermal improvements after installing Thermalright’s LGA1700 contact frame, the other motherboard showed no difference in temperatures whatsoever! Check out our review of this contact frame for more information.

Testing Configuration

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

No Power Limits Cinebench Results

With Raptor Lake’s i9-13900K pushing speeds of 5.5ghz+, even the strongest of coolers hit TJ Max while running Cinebench R23 and other demanding scenarios. As the 13900K is designed to aim for its top safe temperature, we’ll be comparing the overall benchmark score and the CPU’s clock speeds instead. 

The results below are for a 10-minute testing run. But to be sure that was sufficiently long to tax the cooler, we also retested both Thermalright’s Assassin X 120 R SE and DeepCool’s LT720 with a 30-minute Cinebench test. The results didn’t change much at all with the longer test: The average clock speeds maintained dropped by 29mhz on DeepCool’s LT720 and 31mhz on Thermalright’s Assassin X 120 R SE. That’s an incredibly small 0.6% difference in clock speeds maintained, a margin of error difference that tells us that the 10-minute tests are indeed long enough to properly test the coolers.

Because of the difficulty of cooling the i9-13900K in this workload, we feel the best way to compare coolers here is to record the average power consumption of the CPU. When it comes to total cooling capacity, DeepCool’s LT720 is the best-performing unit we’ve tested thus far, cooling an incredible 315W (avg) in this test.

As Cinebench R23’s multi-core benchmark will push coolers to their limits, it’s also a great test for recording the worst-case scenario of fan noise levels. You might notice that our noise charts start at 36dB - this is the noise floor of our testing environment. This makes 36dB our baseline measure, as we’re unable to measure noise levels below this threshold. Keep in mind that noise measurements are logarithmic, meaning the differences between the noise levels of the coolers will be bigger than these graphs would suggest. 

While DeepCool’s LT720 has the best total cooling capacity of the units we’ve tested thus far, it achieves this at the cost of noise. When pushed to full fan speeds, the LT720 is the loudest CPU cooler we’ve recently tested.

We’re also looking at Cinebench R23 benchmark scores to show how little is lost  – or gained – with different cooling solutions. Between the weakest and strongest coolers, there is a total benchmark score variance of 13.7%. 

DeepCool’s LT720’s has the strongest cooling capacity of the coolers we’ve tested thus far with the 13900K, and as a result has the highest Cinebench R23 benchmark score recorded during our testing.

200W Cinebench Results

When restricting power consumption to a more reasonable 200W, DeepCool’s LT720 continues to lead in overall thermal performance, with a 3 degrees Celsius lead over its closest competitor.

Looking at acoustic levels when restricted to 200W, DeepCool’s LT720 sits in the middle of the pack. Not too loud, but not silent either.

125W Cinebench Results

As with our unlimited and 200W testing results, the LT720 continues to hold the crown for best thermal performance in 125W workloads. The lowest power limit I test at is 125W, mainly because this is also the lowest level where I can reliably measure noise measurements. Lower power consumption causes the noise of the CPU cooler to fall below the noise created by the system fans (even while restricted to 35% speed). 

When limited to 125W, the LT720’s fans run very quietly. However, the pump can be heard just barely over the sound of the system fans in this scenario, but this is due to my motherboard’s default settings which encourage higher pump speeds.

Conclusion

The LT720 is a great pairing for demanding CPUs like Intel’s i9-13900K. The refinements of DeepCool’s 4th generation, in-house, water pump design helps the LT720 to take the crown for total cooling capacity - the strongest of any unit I’ve tested thus far, cooling an average of 315W during Cinebench R23 tests. If you need a cooler that can handle the hottest of workloads, DeepCool’s LT720 can handle the heat.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

DeepCool was founded in Beijing in 1996. And the company has, in recent years, delivered some of the best CPU coolers on the market, with options like the Assassin III and AK620. DeepCool’s current lineup includes both air and AIO coolers, as well as computer cases, keyboards, power supplies, and other accessories.   

We most recently reviewed DeepCool’s AK500 cooler, which performed well in our tests. Today we’re going to look at a more budget-oriented product, the single-tower AG400. With the latest CPUs from AMD and Intel requiring ever more capable cooling, is a single tower cooler enough to tame Intel’s 13900K? We’ll have to put it through testing to find out. But first, here are the AG400’s specifications from DeepCool.

Cooler Specifications

Packing and Included Contents 

DeepCool’s AG400 comes in a small box, with molded foam inserts to protect the inner contents.

Included with the package are the following:

• Single-tower heatsink
• 120mm fan
• Fan clips for the included fan
• Mounts for all modern CPU sockets (including AM5 & LGA1700)
• A single-use packet of thermal paste
• Manual

Features of the AXP120-X67

Single Tower Heatsink with Matrix Fin Array

The AG400 (like the larger AK500) incorporates a matrix fin array resembling a checkerboard, which is designed to increase the static pressure of the airflow and therefore better dissipate heat. It also features four 6mm direct-contact copper heatpipes. When installing the AG400, the pre-installed fan does not need to be disassembled during the process, simplifying the installation process and saving time.

Full RAM Clearance

Some coolers with lower-mounted cooling fins won’t fit if you have full-height RAM installed (RGB RAM tends to be taller), but with DeepCool’s AG400 that’s not an issue – it has full RAM height compatibility.

Custom 120mm fan

There’s more to an air cooler than just the heatsink. The bundled fan(s) have a significant impact on cooling and noise levels. Included with the AG400 is an unnamed custom 120mm fan that you won’t find on retail shelves – it’s only available with select DeepCool CPU coolers.

Testing Methodology

While it was fairly easy with previous generations of CPUs for coolers to keep the flagship i9 processor well under TJ max (the maximum temperature a CPU can sustain without throttling) in tough workloads, this is no longer realistically possible on current generation CPUs (the 13900K specifically) without extreme cooling (or enabling power limits).

When I began testing coolers on Intel’s Core i9-12900K, I found that some products that cooled the i9-10900K well struggled in some scenarios when paired with the Alder Lake CPU – and Raptor Lake is even more difficult to cool in these situations. While in the past a CPU hitting its peak temperature was cause for concern, enthusiasts are going to have to learn to accept high temperatures as “normal” while running demanding workloads with Raptor Lake and Ryzen 7000 CPUs.

Modern AMD & Intel CPUs are designed to run fairly hot without any problems– up to 95 degrees C for AMD Ryzen 7000 CPUs, and up to 100 degrees C for Intel’s Core i9-13900K. Similar behavior has been standard in laptops for years due to cooling limitations. 

Furthermore, Intel’s i9-13900K supports Adaptive Boost Technology (ABT) which allows Core i9 processors to dynamically boost to higher all-core frequencies based on available thermal headroom and electrical conditions. This allows multi-core loads to operate at up to 5.5ghz if the necessary amount of thermal dissipation is there. This feature works in a way that actively seeks high temperatures: If the chip sees that it is running underneath the 100-degree C threshold, it will increase its performance and power consumption until it reaches the safe 100C limit, thus sustaining higher clocks (and providing better performance) for longer periods.

The increased cooling challenges posed by Raptor Lake mean that we’ve had to change some of the ways we test coolers. Some coolers were able to pass Cinebench R23 multicore testing with Intel’s 12th Gen i9-12900K when power limits were removed (although only the strongest models were able to pass that test). Most liquid coolers and all air coolers I’ve tested “failed” that test because the CPU reached TJ max in this scenario. 

With Raptor Lake’s 13900K, not a single cooler tested has been able to keep the CPU under TJ max in this test – because as we pointed out, the chip is designed to dial up performance and power until it richest that thermal result. We’ll compare performance instead by looking at total benchmark scores and clock speeds maintained. 

I’ll be testing Intel’s i9-13900K CPU using Asus’ TUF Gaming Z690 Gaming Plus WIFI motherboard and Cooler Master’s HAF 700 Berserker computer case, with case fans limited to 35% speeds. The motherboard’s default fan curve is used for the CPU Cooler’s fans.

In addition to testing Cinebench without power limits enforced, we’ll also be showing results when the CPU’s power consumption is limited to a more reasonable 200W. We’ll also show results at 125W for those who prefer whisper-quiet cooling, at the cost of some performance. For both of these results, we’ll show traditional delta over ambient temperature results.

We’ll provide noise level measurements recorded using a PSPL25 Sound Meter for all three power levels tested, to compare how much noise each cooler makes in different scenarios. We expect most coolers to run effectively silently at 125W.

LGA1700 Socket Bending

Please note there are many factors other than the CPU cooler that can influence your cooling performance, including the case you use and the fans installed in it. A system's motherboard can also influence this, especially if it suffers from bending, which results in poor cooler contact with the CPU.

In order to prevent bending from impacting our cooling results, we’ve installed Thermalright’s LGA 1700 contact frame into our testing rig. If your motherboard is affected by bending, your thermal results will be worse than those shown below. Not all motherboards are affected equally by this issue. I tested Raptor Lake CPUs in two motherboards. And while one of them showed significant thermal improvements after installing Thermalright’s LGA1700 contact frame, the other motherboard showed no difference in temperatures whatsoever! Check out our review of this contact frame for more information.

Testing Configuration

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

No Power Limits Cinebench Results

With Raptor Lake’s i9-13900K pushing speeds of 5.5ghz+, even the strongest of coolers hit TJ Max while running Cinebench R23 and other demanding scenarios. As the 13900K is designed to aim for its top safe temperature, we’ll be comparing the overall benchmark score and the CPU’s clock speeds instead. 

The results below are for a 10-minute testing run. But to be sure that was sufficiently long to tax the cooler, we also retested both Thermalright’s Assassin X 120 R SE and DeepCool’s LT720 with a 30-minute Cinebench test. The results didn’t change much at all with the longer test: The average clock speeds maintained dropped by 29mhz on DeepCool’s LT720 and 31mhz on Thermalright’s Assassin X 120 R SE. Looking at this another way, that’s an incredibly small 0.6% difference in clock speeds maintained, a margin of error difference that tells us that the 10 minute tests are indeed long enough to properly test the coolers.

Because of the difficulty of cooling the i9-13900K in this workload, we feel the best way to compare coolers here is to record the average power consumption of the CPU. DeepCool’s AG400 does alright here, cooling about 247W on average, placing it in the middle of the pack as far as performance goes (in comparison to the other coolers tested with the new CPU so far).

We’re also showing the total Cinebench R23 benchmark scores to show how little is lost  – or gained – with different cooling solutions. Between the weakest and strongest coolers, we have total benchmark score variance of roughly 13%. DeepCool’s AG400 sits right in the middle of the scores again, providing 94% of the benchmark score of the strongest AIO we’ve tested thus far.

As Cinebench R23’s multi-core benchmark will push coolers to their limits, it’s also a great test for recording the worst-case scenario of fan noise levels. Here, DeepCool’s AG400 ran a bit louder than most other coolers we’ve tested with Raptor Lake, but quieter than the IceSLEET G6 Stealth and DeepCool’s LT720 AIO.

200W Cinebench Results

When restricting power consumption to a more reasonable 200W, the AG400 is basically tied with Thermalright’s Assassin 120 R SE and the IceSLEET G6 Stealth.

Looking at acoustic levels here, the AG 400’s performance is still on the noisier side of things. It’s certainly quieter than Iceberg Thermal’s IceSLEET G6 Stealth, but significantly louder than Thermalright’s Assassin 120 R SE.

125W Cinebench Results

The lowest power limit I test at is 125W, mainly because this is also the lowest level where I can reliably measure noise measurements. Lower power consumption causes the noise of the CPU cooler to fall below the noise created by the system fans (even while restricted to 35% speed).

When limited to 125W, DeepCool’s AG400 runs silently and is essentially indistinguishable from the rest of the system fans. It does run a bit warmer than most other coolers tested here, but this isn’t a cause for concern given that the increased temperature here doesn’t impact performance.

Conclusion

At only $30, DeepCool’s AG400 offers good value and decent performance – even when paired with CPUs like Intel’s i9-13900K. While it does get a bit louder than other coolers in the heaviest workloads, in less thermally taxing scenarios (which should be the vast majority of the time) the AG400 runs whisper quiet and makes a great pairing with any modern CPU.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

DeepCool was founded in Beijing in 1996 and has earned a solid reputation in the CPU Cooler space with options like the Assassin III and AK620, considered to be some of the best air coolers on the market. The company’s lineup includes both air and AIO coolers as well as computer cases, keyboards, power supplies and other accessories.  

While most AIOs on the market are based on Asetek designs, DeepCool is one of the few companies which makes its own, in-house, pump designs. DeepCool sent us the new LS320, a 120mm AIO liquid cooler featuring its latest generation in-house design, and the cooler retails for $89.99. 

In the past I’ve generally avoided 120mm AIOs due to their typically limited cooling potential. However, we previously tested the larger 240mm LS520 cooler and were impressed by how well it handled Intel’s Core i9-12900K. That made me wonder: Is the LS320’s redesigned pump system enough to overcome the disadvantages of a smaller radiator and earn a spot on our best AIO coolers list? We’ll have to put it through testing to find out. But first here are the specifications from DeepCool. 

Specifications for the DeepCool LS520

Packing and Included Contents 

DeepCool’s LS320 ships in a small, almost square box, packed with molded cardboard and soft plastic coverings of the individual parts for protection. 

Included with the package are the following:

• CPU Block and Radiator
• 1x FC120 120mm fan
• Mounts for all modern CPU sockets, including LGA1700, TR4, and AM4/AM5 motherboards
• Customizable blank plate for CPU block
• Manual & support pamphlets
• 2x coolant line clips

Cooler Installation

Installing the DeepCool’s LS320 was fairly easy. To begin, you’ll first want to secure the radiator to your case. And with these small 120mm AIOs, the radiator often gets mounted in the case’s rear exhaust, though we mounted at the top for comparison to larger coolers. For more about radiator and fan placement, see our PC airflow guide. After the radiator and fan are mounted, press the backplate against the motherboard and mount the stand-offs. You won’t need to worry about thermal paste, as it is pre-applied to the CPU block here. Press the CPU block against the standoffs, and then use the included thumb screws to secure it.

New Testing Configuration

What's different than other coolers?

Latest in-house pump design from DeepCool 

The vast majority of Liquid Coolers on the market today are based around Asetek designs, with pumps integrated into the CPU block. But the LS520 incorporates the latest generation in-house design from DeepCool.

User Customizable Plate  

Included with the LS320 is a blank CPU block plate, which lets you create your own design for the RGB-lit CPU block. Given that most companies just slap their logo here and give you no other options, this is a nice touch.

Custom DeepCool FC120 fan

The LS320 includes DeepCool’s FC120 fan, sporting a unique fan connection system which allows for easy daisy-chaining of both fan & lighting controls in a single cable. The FC120 fans are a stronger SKU than the company’s retail models and feature upgraded airflow and static pressure as a result of higher maximum fan speeds. These fans include arrows on the sides indicating both the directions that the fans spin as well as the direction of the airflow.

As with other recent AIOs, I'll be testing DeepCool’s LS320 with Intel's Core i9-12900K. Due to the increased thermal density of the Intel 7 manufacturing process, as well as changes to core and component layouts, Alder Lake CPUs are more difficult to cool than previous generation CPUs in the most heat-intensive of workloads. 

This means that coolers that kept previous generation products like the i9-10900K nice and cool sometimes struggle to keep Intel's i9-12900K under TJ max–the top temperature before the CPU starts to throttle. Many coolers, including DeepCool’s last generation Castle EX AIOs, failed to keep the i9-12900K under TJ max when power limits are removed in workloads like Cinebench and OCCT when I’ve tested them.

Please note there are many factors other than the CPU cooler that can influence your cooling performance. A system's motherboard can especially influence this, as there are boards on the market with CPU sockets that aren't up to Intel's spec, which can cause warping or poor contact with the CPU. The case you use will also influence cooling results.

With Alder Lake's cooling demands in mind, I'll be rating CPU Coolers in 3 different tiers.

Tier 1: These coolers are able to keep the i9-12900K below TJ max in most loads, with no power limits enforced. I expect only the best liquid coolers to meet this standard. 

Tier 2: These coolers are able to keep the i9-12900K under the TJ max threshold with CPU power limits of 200W enforced. I expect most liquid coolers and the best air coolers to meet this standard.

Tier 3: These coolers are able to keep the i9-12900K under TJ max with CPU power limits of 140W enforced.

Testing Methodology

To test the limits of a cooler's thermal dissipation capabilities, I run two primary stress tests: Cinebench and OCCT each for 10 minutes. While this may be a short amount of time, it is sufficient to push most coolers–air and liquid–to their limits. 

While stress testing in Cinebench, I run both with power limits removed and with an enforced 200W CPU power limit. In this test setup using MSI’s z690 A Pro DDR4 Motherboard and Be Quiet’s Silent Base 802 Computer Case. Only the best coolers are able to pass Cinebench testing when power limits are removed. 

I don’t test OCCT without power limits because attempting to do so results in CPU package power consumption jumping to over 270W and instantly throttling with even the best AIO coolers. Instead, I test at 200W to give coolers a chance at passing. I also include 140W results to give data closer to a lower-end CPU, such as AMD’s Ryzen 5600X or Intel’s i5-12400.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

Cinebench Test Results

In the past, I’ve begun my cooler testing with Intel’s i9-12900K by testing Cinebench without power limits and with a 200W power limit enforced. Unfortunately, DeepCool’s LS320 wasn’t capable of passing either of these tests. However, this is to be expected as it’s a 120mm AIO and users who want a cooler capable of cooling higher TDP loads should be looking at 240/280mm sized AIOs or larger.

200W and 140W OCCT Test Results

I usually like to run OCCT's small set stress testing for stability when overclocking, but on Alder Lake I haven't found a cooler that's capable of handling OCCT without throttling unless power limits are enforced.

I test OCCT at 200W to demonstrate a thermally demanding load, but also with a 140W power limit enforced to show how these coolers might perform with a CPU that's easier to cool, like Intel's i5-12600K or AMD's Ryzen 5800X.

Just like with Cinebench, the LS320 wasn’t able to pass our 200W testing here. However, when restricted to a more reasonable 140W, the LS320 handled our cooling tests easily. It performed about equal to the Corsair H100i Elite and Cooler Master’s PL360 Flux when limited to 140W and 50% fan speeds.

95W OCCT Test Results

I’ve begun to include 95w testing results due to reader requests. These results should be representative of performance on CPUs like AMD’s Ryzen 5 5600X or Intel’s i5-12400. In this testing, the LS320 performed well–on par with most coolers tested at this power level.

Noise Levels and Acoustics

To test noise levels, I used the SLM25TK Sound Level Meter positioned 18 inches behind the rear of the Be Quiet Silent Base 802 PC case, and recorded early in the morning to achieve the lowest noise floor possible. The chart below shows averaged results, measured over the course of five minutes, to account for sudden variations in measurements.

The LS320 runs fairly quietly in workloads it was designed to handle, running realistic workloads with TDPs of 140W or less will result in fairly quiet operation. This cooler output only 34.5 dBA when restricted to 50% fan speeds. If you attempt to cool high TDP loads on it, noise levels will be comparable to Be Quiet’s Pure Loop 2 FX or Cooler Master’s PL240 Flux, as the fan hits 100% RPM speeds.

Conclusion

DeepCool’s LS320 is a small AIO, which is best paired with CPUs of mid-range TDPs, like Intel’s i5-12600K. It’s capable of handling 140W-plus when paired with Alder Lake CPUs, although it will get noisy as the fan speed maxes out. The LS320 is a bit expensive for the performance it provides, but it will work well in SFF cases that don’t have room for bigger coolers.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

DeepCool was founded in Beijing in 1996 and has earned a solid reputation in the CPU Cooler space with options like the Assassin III and AK620, considered to be some of the best air coolers on the market. The company’s lineup includes both air and AIO coolers as well as computer cases, keyboards, power supplies, and other accessories.  

We have DeepCool’s new LS520, a 240mm AIO liquid cooler which retails for $109 USD, on our test bench. While most AIOs on the market are based on Asetek designs, DeepCool is one of the few companies which make their own, in-house, pump designs. The LS520 features a redesigned pump system, but is that enough to tame Intel’s 12900K and earn a spot on our best AIO coolers list? We’ll have to put it through testing to find out. But first here are the specifications from DeepCool. 

Specifications for the DeepCool LS520

Packing and Included Contents

DeepCool’s LS520 ships in a medium-sized box, packed with molded cardboard and soft p;lastic coverings of the individual parts for protection. 

Included with the package are the following:

• CPU Block + Radiator
• 2x FC120 120mm fans
• Mounts for all modern CPU sockets, including LGA1700, TR4, and AM4/AM5 motherboards
• Customizeable blank plate for CPU Block
• User Manual & Support Pamphlets
• Coolant Line Clips

Cooler Installation

Installing the DeepCool’s LS520 was fairly easy. To begin, you’ll first want to secure the radiator to your case. Afterwards, press the backplate against the motherboard and mount the stand-offs. You won’t need to worry about thermal paste, as it is pre-applied to the CPU block. Press the CPU block against the standoffs, and then use the included thumb screws to secure it. 

New Testing Configuration

What's different than other coolers?

Latest in-house pump design from DeepCool 

The vast majority of Liquid Coolers on the market today are based around Asetek designs, with pumps integrated into the CPU block. But the LS520 incorporates the latest generation in-house design from DeepCool.

User Customizable Plate

Included with the LS520 is a blank CPU block plate, which lets you create your own design for the RGB-lit CPU block. Given that most companies just slap their logo here and give you no other options, this is a nice touch. 

Custom DeepCool FC120 fans

The LS520 includesDeepCool’s FC120 fans, sporting a unique fan connection system which allows for easy daisy-chaining of both fan & lighting controls in a single cable. The FC120 fans are a stronger SKU than the retail models and feature upgraded airflow and static pressure as a result of higher maximum fan speeds.

Testing Configuration

I'll be testing DeepCool’s LS520 with Intel's i9-12900K. Due to the increased thermal density of the Intel 7 manufacturing process, as well as changes to core and component layouts, Alder Lake CPUs are more difficult to cool than previous generation CPUs in the most heat-intensive of workloads. 

This means that coolers that kept previous generation products like the i9-10900K nice and cool sometimes struggle to keep Intel's i9-12900K under Tj max–the top temperature before the CPU starts to throttle. Many coolers, including DeepCool’s last generation CASTLE EX AIOs, failed to keep the i9-12900K under TJ max when power limits are removed in workloads like Cinebench and OCCT when I’ve tested them.

Please note there are many factors other than the CPU cooler that can influence your cooling performance. A system's motherboard can especially influence this, as there are boards on the market with CPU sockets that aren't up to Intel's spec, which can cause warping or poor contact with the CPU. The case you use will also influence cooling results.

With Alder Lake's cooling demands in mind, I'll be rating CPU Coolers in 3 different tiers.

Tier 1: These coolers are able to keep the i9-12900K below TJ max in most loads, with no power limits enforced. I expect only the best liquid coolers to meet this standard. 

Tier 2: These coolers are able to keep the i9-12900K under the TJ max threshold with CPU power limits of 200W enforced. I expect most liquid coolers and the best air coolers to meet this standard. 

Tier 3: These coolers are able to keep the i9-12900K under TJ max with CPU power limits of 140W enforced.

Testing Methodology

To test the limits of a cooler's thermal dissipation capabilities, I run two primary stress tests: Cinebench and OCCT each for 10 minutes. While this may be a short amount of time, it is sufficient to push most coolers–air and liquid–to their limits. 

While stress testing in Cinebench, I run both with power limits removed and with an enforced 200W CPU power limit. In this test setup using MSI’s z690 A Pro DDR4 Motherboard and Be Quiet’s Silent Base 802 Computer Case. Only the best coolers are able to pass Cinebench testing when power limits are removed. 

I don’t test OCCT without power limits because attempting to do so results in CPU package power consumption jumping to over 270W and instantly throttling with even the best AIO coolers. Instead, I test at 200W to give coolers a chance at passing. I also include 140W results to give data closer to a lower-end CPU, such as AMD’s Ryzen 5600X or Intel’s i5-12400.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

Cinebench Test Results

Running Cinebench without power limits is the most thermally demanding load in my testing, and most coolers don't pass this specific test–but the LS520 isn’t most coolers. It performed exceptionally well here, beating BeQuiet’s Pure Loop 2 FX 360mm by 11 degrees Celsius!  

When power limits are restricted to a reasonable 200W and 50% fan speeds, the LS520 ties with Cougar’s Poseidon AIO and is close to Be Quiet’s Pure Loop 2 FX. And keep in mind these are both coolers with larger 36mm radiators.

OCCT Test Results

I usually like to run OCCT's small set stress testing for stability when overclocking, but on Alder Lake I haven't found a cooler that's capable of handling OCCT without throttling unless power limits are enforced.

I test OCCT at 200W to demonstrate a thermally demanding load, but also with a 140W power limit enforced to show how these coolers might perform with a CPU that's easier to cool, like Intel's i5-12600k or AMD's Ryzen 5800X.

OCCT’s cooling results are similar to the previous results when limited to 200W. The LS520 nearly ties with Cougar’s Poseidon GT630 here, and cools slightly better than BeQuiet’s Pure Loop 2 FX.

Noise Levels and Acoustics

To test noise levels, I used the SLM25TK Sound Level Meter positioned 18 inches behind the rear of the Be Quiet Silent Base 802 PC case, and recorded early in the morning to achieve the lowest noise floor possible. The chart below shows averaged results, measured over the course of five minutes, to account for sudden variations in measurements.

Even while pushing over 230w, DeepCool’s LS520 doesn’t get very loud. At 50% fan speeds it’s a little louder than the BeQuiet & Cougar models, and at full speeds it sits between them in noise levels. 

Conclusion

With Alder Lake’s increased cooling demands, many coolers can’t handle the heat. DeepCool’s LS520 packs incredible cooling performance in a small package, able to handle Alder Lake without power limits in most situations. This is the best performing 240mm AIO I have tested thus far, and at only $109 its value is unbeatable.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

DeepCool was founded in Beijing in 1996 and has, in recent years, earned a solid reputation in the CPU Cooler space with options like the Assassin III and AK620, considered to be some of the best CPU air coolers on the market. The company’s lineup includes both air and AIO coolers, as well as computer cases, keyboards, power supplies, and other accessories.  

But with the latest CPUs from AMD and Intel requiring ever more capable cooling, DeepCool is back with a new AK500 air cooler. The AK500 features a chungus radiator and a 120mm FK120 fan, but is that enough to tame Intel’s 12900K and earn a spot on our best coolers list? We’ll have to put it through testing to find out. But first here are the AK50’s specifications from DeepCool.

Specifications for the DeepCool AK500 

Packing and Included Contents

DeepCool’s AK500 is packaged in a medium-sized box, with cardboard and foam to protect the cooler during shipping.

Included with the package are the following:

• Thick heatsink
• 1x DeepCool FK120 fan
• Mounts for all modern CPU sockets, including AM5 & LGA1700
• A full-sized tube of thermal paste
• Manual
• Screwdriver
• Low-speed fan adapter
• Clips for a second fan (not included)

Cooler Installation

Installing DeepCool’s AK500 is mostly easy. To begin, you’ll press the backplate against your motherboard and then use the thumbscrew standoffs to secure it. Afterward, set the mounting bars on the standoffs and secure them using the included thumbscrews.

The AK500’s thermal paste isn’t pre-installed, but DeepCool instead includes a medium-sized tube of thermal paste that can be used for multiple applications. After applying the thermal paste, you’ll need to secure the heatsink against the mounting bars with the captive screws. This is where it can be a little tricky. The screw on the right is easy enough to secure, but the screw on the left requires the use of a long screwdriver through the body of the heatsink, and it can be tricky to aim properly.

New Testing Configuration

What's different than other coolers?

Angled Heatpipes with tall DRAM compatibility

The heatpipes on the AK500 have an angled offset, which allows for compatibility with tall RAM – you shouldn’t have to worry about RAM height causing installation problems. DeepCool says the internal capillary structure of the heatpipes has been optimized to achieve high heat dissipation, no matter the orientation of the cooler, whether installed vertically or horizontally.

Optimized Metal Mounting Bracket

The included metal mounting bracket has been optimized to ensure proper contact on the latest Intel and AMD platforms.

Mounting Clips for a 2nd fan

While the AK500 only includes one fan, DeepCool has included an extra pair of fan clips for enthusiasts who want to add a second fan for maximum cooling performance.

1x DeepCool FK120 fan

There’s more to an air cooler than just a heatsink. The fan included has a significant impact on cooling, noise levels, and performance. Included with DeepCool’s AK500 is a single 120mm FK120 model fan. On the sides of the LK120 fans, there are two arrows. One arrow shows the direction the fans spin, the other indicates the direction of the airflow.

Checkerboard Matrix Fin Array

The AK500 incorporates a matrix fin array, resembling a checkerboard, which is designed to increase the static pressure of the airflow.

Testing Configuration

I'll be testing DeepCool’s AK500 paired with Intel’s i9-12900K CPU installed in MSI’s z690 A PRO DDR4 motherboard and BeQuiet’s Silent Base 802 Computer Case. I’ve mentioned in previous reviews that due to the increased thermal density of the Intel 7 manufacturing process, as well as changes to core and component layouts, Alder Lake CPUs are more difficult to cool than previous generation CPUs in the most core-intensive of workloads.

This means that coolers that kept previous generation products like the i9-10900K nice and cool sometimes struggle to keep Intel's i9-12900K under Tj max (the top temperature before the CPU starts to throttle). Many coolers I’ve tested have failed to keep the i9-12900K under TJ max when power limits are removed in workloads like Cinebench and OCCT.

For this review, I’ll be primarily comparing DeepCool’s AK500 to Cougar’s Forza 85, but I’ll also include results from various liquid coolers I’ve tested recently.

Please note there are many factors other than the CPU cooler that can influence your cooling performance. A system's motherboard can especially influence this, as there are boards on the market with CPU sockets that aren't up to Intel's spec, which can cause warping or poor contact with the CPU. The case you use will also influence cooling results.

With Alder Lake's cooling demands in mind, I'll be rating CPU Coolers in 3 different tiers.

Tier 1: These coolers are able to keep the i9-12900K below TJ max in most loads, with no power limits enforced. I expect only the best liquid coolers to meet this standard.

Tier 2: These coolers are able to keep the i9-12900K under the TJ max threshold with CPU power limits of 200W enforced. I expect most liquid coolers and the best air coolers to meet this standard.

Tier 3: These coolers are able to keep the i9-12900K under TJ max with CPU power limits of 140W enforced.

Testing Methodology

To test the limits of a cooler's thermal dissipation capabilities, I run two primary stress tests: Cinebench and OCCT each for 10 minutes. While this may be a short amount of time, it is sufficient to push most coolers – air and liquid – to their limits.

While stress testing in Cinebench, I run both with power limits removed and with an enforced 200W CPU power limit. In this test setup using MSI’s z690 A Pro DDR4 Motherboard and Be Quiet’s Silent Base 802 Computer Case, only the best coolers are able to pass Cinebench testing when power limits are removed.

I don’t test OCCT without power limits because attempting to do so results in CPU package power consumption jumping to over 270W and instantly throttling with even the best AIO coolers. Instead, I test at 200W to give coolers a chance at passing. I’ve included 140W results to give data comparable to a CPU that doesn’t use as much power, such as AMD’s Ryzen 5800X or Intel’s i5-12600K. 95W test results are also included, for those interested in results comparable to low TDP CPUs such as AMD’s Ryzen 5600x or Intel’s i5-12400.

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

Cinebench Test Results

I first attempted to run Cinebench without power limits enforced. If the cooler can handle it, this load will use a little more than 230W. Unfortunately, I haven’t found an air cooler capable of passing this test, and DeepCool’s AK500 is no exception.

However, when testing Cinebench limited to a more reasonable 200W CPU power limit, DeepCool’s AK500 averaged 68 degrees Celcius over ambient, an impressive 5 degrees C cooler than Cougar’s Forza 85 Air Cooler. The AK500 wasn’t able to pass 200W testing when fan speeds were reduced to 50%, but it was able to handle 175W loads with reduced fan speeds.

OCCT Test Results

I usually like to run OCCT's small set stress testing for stability when overclocking, but on Alder Lake I haven't found a cooler that's capable of passing unless power limits are enforced. This scenario pulls over 270W and causes even the best coolers to throttle.

For Alder Lake, I prefer to test OCCT at 200W to demonstrate a thermally demanding load, but also with a 140W power limit enforced to show how these coolers might perform with a CPU that's easier to cool, like Intel's i5-12600K or AMD's Ryzen 5800X.

While running the OCCT stress testing at both 200W, DeepCool’s AK500 outperformed the Cougar Forza 85 by an impressive 4 degrees Celcius. When limited to a CPU power limit of 140w, these coolers performed on par with each other – except when fans were limited to 50% speeds, where the AK500 showed a slight lead over the Forza 85.

Noise Levels and Acoustics

To test noise levels, I used the SLM25TK Sound Level Meter positioned 18 inches behind the rear of the Be Quiet Silent Base 802 PC case, and recorded early in the morning to achieve the lowest noise floor possible. The chart below shows averaged results, measured over the course of ten minutes, to account for sudden variations in measurements.

Even in the hottest of workloads with the FK120 fan running at full speed (above 200W of CPU package power), DeepCool’s AK500 stays extremely silent. The full-speed fan noise levels are the second quietest of all coolers I’ve tested on this platform thus far, and the AK500’s half-speed noise levels are quieter than all but two coolers we have tested thus far.

Conclusion

DeepCool’s AK500 performs great when paired with Alder Lake CPUs like the i9-12900K, and can handle loads of over 200W – and it keeps the CPU cool while running nice and quiet. With an MSRP of just $59.99, the AK500 delivers an impressive combination of silence, performance, and value that is hard to beat!

• MORE: How to Buy the Right CPU Cooler
• MORE: How to Overclock a CPU
• MORE: How to Check CPU Temperature
• MORE: All CPU Cooling Content

The Deepcool PQ850M uses a pretty good and reliable Seasonic platform, which has started to show its age, though. Thanks to its terrific CWT platform, the Corsair RM850x (2021) remains the performance king in the 850W Gold category. The weird exterior design of the PQ850 doesn't help, either. Another decent option with similar specs from Deepcool is the DQ850-M-V2L, which achieves better performance than the PQ850M, using a budget CWT platform. A revision of the PQ850M is required, featuring a 12+4 pin connector, to give it a chance for inclusion in our best PSUs article. 

Deepcool enriched its PSU lines with the PQ-M series, including four models with 650W to 1000W max power. All are fully modular and use the trendy Seasonic Focus Gold platform. They have compact dimensions and are equipped with a 120mm diameter fluid dynamic bearing fan provided by Hong Hua. The exterior design looks weird because of the square holes of the fan grille. The design definitely stands out from the crowd, but not in a good way, according to our taste, at least. 

This review will evaluate the second strongest member of the PQ-M line, the 850W unit, which is strong enough to support a capable gaming system. The PSU doesn't have a 12+4 pin PCIe connector, meaning that it isn't compatible with either the newest ATX spec (v3.0) or upcoming GPUs. It is not ideal to buy a PSU right now because we are in a transition phase from the previous ATX spec to the new one. 

The only graphics card, so far at least, that comes with a 12+4 pin PCIe connector is the Nvidia RTX 3090 Ti, and it has an adapter from three 6+2 PCIe to a single 12+4 PCIe. Still, you need the PSU to be as future-proof as possible, and the upcoming GPUs will utilize the extra 4-pins so that adapters won't be fully compatible. 

An extended, ten-year warranty covers the PQ850M, and its MSRP is set at 169 dollars, which is a fair amount given its features. It is certified as Gold by 80 PLUS and Platinum by Cybenetics. It also has a noise rating of Cybenetics A- (25-30 dB[A]). 

Specifications

Power Specifications

Cables & Connectors

The only three PCIe connectors are on dedicated cables, so this PSU can even power an RTX 3090 Ti, using the provided adapter. Given that it is not wise to use PCIe cables with more than one connector and that SLI and Crossfire are long dead, there is no need for more than three PCIe connectors on dedicated cables in an 850W PSU. 

There are no in-cable caps, and all cables are long enough. If only the distance between the peripheral connectors was longer, at 150mm, at least. Finally, it would be nice to see thicker gauges on the ATX, EPS, and PCIe connectors. 

Component Analysis

We strongly encourage you to have a look at our PSUs 101 article, which provides valuable information about PSUs and their operation, allowing you to better understand the components we're about to discuss.

Deepcool utilized Seasonic's trendy Focus Plus Gold design for this series. This is a good and reliable platform but not fully compatible with the newest ATX spec. The build quality is good, and as usual, Seasonic used good parts everywhere. The heat sinks are small, and there is ample space between parts for good airflow, so the Hong Hua FDB fan won't have to work over hours. 

The transient/EMI filter has all necessary parts, including an MOV and a discharge IC, providing a small efficiency boost. There is also an NTC thermistor for lowering inrush currents, supported by a bypass relay. 

The pair of bridge rectifiers can handle up to 30A of current. 

The APFC converter uses two Infineon FETs and a single boost diode provided by STMicronics. The bulk cap is by Chemi-Con, and it has enough capacity to provide a longer than 17ms hold-up time. The APFC controller is a Champion CM6500UNX. 

The main FETs are four Great Power, installed in a full-bridge topology. An LLC resonant converter is also used to increase efficiency. The resonant controller is a Champion CM6901T6.

Four Nexperia FETs regulate the 12V rail. Two DC-DC converters generate the minor rails. 

The electrolytic filtering caps are by Chemi-Con so that they will outlive the provided warranty. Lots of polymer caps are also used. 

The standby PWM controller is an Excelliance MOS EM8569C. 

The modular board has many polymer and two electrolytic caps. 

The main supervisor IC is a Weltrend WT7527V.

Soldering quality is good. 

The cooling fan measures 120mm across and it uses a fluid dynamic bearing. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

To learn more about our PSU tests and methodology, please check out How We Test Power Supply Units. 

Primary Rails And 5VSB Load Regulation

The following charts show the main rails' voltage values recorded between a range of 40W up to the PSU's maximum specified load, along with the deviation (in percent). Tight regulation is an important consideration every time we review a power supply because it facilitates constant voltage levels despite varying loads. Tight load regulation also, among other factors, improves the system’s stability, especially under overclocked conditions and, at the same time, it applies less stress to the DC-DC converters that many system components utilize.

Load regulation is perfect on the main three rails and acceptable at 5VSB, where it doesn't matter much as long as this rail keeps its voltages within the specified range by the ATX spec. 

Hold-Up Time

Put simply; hold-up time is the amount of time that the system can continue to run without shutting down or rebooting during a power interruption.

The hold-up time is longer than 17ms, but the power ok signal is lower than 16ms. 

Inrush Current

Inrush current, or switch-on surge, refers to the maximum, instantaneous input current drawn by an electrical device when it is first turned on. A large enough inrush current can cause circuit breakers and fuses to trip. It can also damage switches, relays, and bridge rectifiers. As a result, the lower the inrush current of a PSU right as it is turned on, the better.

Inrush current is low with 115V but pretty high with 230V. A higher resistance NTC thermistor would help here. 

Leakage Current

In layman's terms, leakage current is the unwanted transfer of energy from one circuit to another. In power supplies, it is the current flowing from the primary side to the ground or the chassis, which in the majority of cases is connected to the ground. For measuring leakage current, we use a GW Instek GPT-9904 electrical safety tester instrument.

The leakage current test is conducted at 110% of the DUT's rated voltage input (so for a 230-240V device, we should conduct the test with 253-264V input). The maximum acceptable limit of a leakage current is 3.5 mA and it is defined by the IEC-60950-1 regulation, ensuring that the current is low and will not harm any person coming in contact with the power supply's chassis.

Leakage current is higher than competing offerings, but still lower than the limit. 

10-110% Load Tests

These tests reveal the PSU's load regulation and efficiency levels under high ambient temperatures. They also show how the fan speed profile behaves under increased operating temperatures.

The PSU doesn't sweat at high operating temperatures, but its fan gets screaming loud. Moreover, efficiency takes a huge hit, so you better not push it so hard. 

20-80W Load Tests

In the following tests, we measure the PSU's efficiency at loads significantly lower than 10% of its maximum capacity (the lowest load the 80 PLUS standard measures). This is important for representing when a PC is idle with power-saving features turned on.

The fan is not needed to spin under light loads, even under increased temperatures. 

2% or 10W Load Test

From July 2020, the ATX spec requires 70% and higher efficiency with 115V input. The applied load is only 10W for PSUs with 500W and lower capacities, while for stronger units, we dial 2% of their max-rated capacity.

It would be nice to see over 70% efficiency with a 2% load. 

Efficiency & Power Factor

Next, we plotted a chart showing the PSU's efficiency at low loads and loads from 10 to 110% of its maximum rated capacity. The higher a PSU’s efficiency, the less energy goes wasted, leading to a reduced carbon footprint and lower electricity bills. The same goes for Power Factor.

With normal loads, the competition is ahead. With 20-80W loads, the PQ850M performs better. Finally, with a 2% load, it cannot reach 70% efficiency, which is not a problem for the EVGA 850 G6, which uses a more advanced version of the Focus platform.

5VSB Efficiency

We would like to see over 79% average efficiency at 5VSB. 

Power Consumption In Idle And Standby

Vampire power is low.

Fan RPM, Delta Temperature, And Output Noise

All results are obtained between an ambient temperature of 37 to 47 degrees Celsius (98.6 to 116.6 degrees Fahrenheit).

The fan speed profile pushes the small fan to its limits, under high operating conditions. 

The following results were obtained at 30 to 32 degrees Celsius (86 to 89.6 degrees Fahrenheit) ambient temperature.       

At normal operating temperatures, close to 30 degrees Celsius, the PSU is silent with up to 620W loads. It exceeds 30 dBA with more than 690W, and it enters the 35-40 dBA zone with more than 750W at 12V. All in all, this is not a noisy PSU. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Protection Features

Check out our PSUs 101 article to learn more about PSU protection features.

The OCP triggering points at low temperatures are higher than usual, but this doesn't affect the PSU's operation. With increased temperatures, OCP at 12V drops notably lower, protecting the platform more effectively. The same goes for OPP. At low temperatures, it is pretty high, while with increased ambient, it drops 200W lower. We don't usually see so much difference in OCP and OPP under cold and hot conditions. 

DC Power Sequencing

According to Intel’s most recent Power Supply Design Guide (revision 1.4), the +12V and 5V outputs must be equal to or greater than the 3.3V rail at all times. Unfortunately, Intel doesn't mention why it is so important to always keep the 3.3V rail's voltage lower than the levels of the other two outputs.

No problems here since the 3.3V rail is always lower than the other two. 

Cross Load Tests

To generate the following charts, we set our loaders to auto mode through custom-made software before trying more than 25,000 possible load combinations with the +12V, 5V, and 3.3V rails. The deviations in each of the charts below are calculated by taking the nominal values of the rails (12V, 5V, and 3.3V) as point zero. The ambient temperature during testing was between 30 to 32 degrees Celsius (86 to 89.6 degrees Fahrenheit).

Load Regulation Charts

Efficiency Graph

Ripple Graphs

The lower the power supply's ripple, the more stable the system will be and less stress will also be applied to its components.

Infrared Images

We apply a half-load for 10 minutes with the PSU's top cover and cooling fan removed before taking photos with a modified Fluke Ti480 PRO camera able to deliver an IR resolution of 640x480 (307,200 pixels).

Without active cooling, the boost diode gets hot, very hot. This is strange since it is firmly secured on the APFC heat sink, using a thermal pad for better contact. Nonetheless, this diode can handle 8A with 125 degrees Celsius junction temperature. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Advanced Transient Response Tests

For details about our transient response testing, please click here.

In the real world, power supplies are always working with loads that change. It's of immense importance, then, for the PSU to keep its rails within the ATX specification's defined ranges. The smaller the deviations, the more stable your PC will be with less stress applied to its components. 

We should note that the ATX spec requires capacitive loading during the transient rests, but in our methodology, we also choose to apply a worst case scenario with no additional capacitance on the rails. 

Advanced Transient Response at 20% – 20ms

Advanced Transient Response at 20% – 10ms

Advanced Transient Response at 20% – 1ms

Advanced Transient Response at 50% – 20ms

Advanced Transient Response at 50% – 10ms

Advanced Transient Response at 50% – 1ms

Transient response is good enough at 12V and 5VSB, decent at 5V and mediocre at 3.3V.

Turn-On Transient Tests

In the next set of tests, we measure the PSU's response in simpler transient load scenarios—during its power-on phase. Ideally, we don't want to see any voltage overshoots or spikes since those put a lot of stress on the DC-DC converters of installed components.

There is a tiny voltage overshoot at 5VSB, while at 12V, we meet the usual steps and some rigging in the last test, where we apply a full load immediately after switching on the PSU. 

Power Supply Timing Tests

There are several signals generated by the power supply, which need to be within specified, by the ATX spec, ranges. If they are not, there can be compatibility issues with other system parts, especially mainboards. From year 2020, the PSU's Power-on time (T1) has to be lower than 150ms and the PWR_OK delay (T3) from 100 to 150ms, to be compatible with the Alternative Sleep Mode.

The PWR_OK delay is way above the 100-150ms region, so the PSU does not support the alternative sleep mode recommended by the ATX spec. The power-on time is below 100ms, at least, but it could be close to 50ms, to not create compatibility issues with "picky" mainboards. 

Ripple Measurements

Ripple represents the AC fluctuations (periodic) and noise (random) found in the PSU's DC rails. This phenomenon significantly decreases the capacitors' lifespan because it causes them to run hotter. A 10-degree Celsius increase can cut into a cap's useful life by 50%. Ripple also plays an important role in overall system stability, especially when overclocking is involved.

The ripple limits, according to the ATX specification, are 120mV (+12V) and 50mV (5V, 3.3V, and 5VSB).

There is a significant increase in the 12V rail's ripple from 90% to full load, which doesn't look nice on the charts. Still, you cannot call ripple suppression bad since 33.7mV at 12V under sky-high operating temperatures is not that high. 

Ripple At Full Load

Ripple At 110% Load

Ripple At Cross-Load 1

Ripple At Cross-Load 4

EMC Pre-Compliance Testing – Average & Quasi-Peak EMI Detector Results

Electromagnetic Compatibility (EMC) is the ability of a device to operate properly in its environment without disrupting the proper operation of other nearby devices.

Electromagnetic Interference (EMI) stands for the electromagnetic energy a device emits, and it can cause problems in other nearby devices if too high. For example, it can cause increased static noise in your headphones or/and speakers.

΅We use TekBox's EMCview to conduct our EMC pre-compliance testing. 

Conducted EMI emissions are low. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Performance Rating

The overall performance cannot meet the competition. The difference is not huge, though, but the second from the bottom place still doesn't look good. 

Noise Rating

The graph below depicts the cooling fan's average noise over the PSU's operating range, with an ambient temperature between 30 to 32 degrees Celsius (86 to 89.6 degrees Fahrenheit).

Under normal operating temperatures, the average noise output is low and this is a major asset for the PQ850M

Efficiency Rating

The following graph shows the PSU's average efficiency throughout its operating range with an ambient temperature close to 30 degrees Celsius.

The average efficiency score is decent. It would be nice to be closer to or above 89%. 

Power Factor Rating

The following graphs show the PSU's average power factor reading throughout its operating range with an ambient temperature close to 30 degrees Celsius and 115V/230V voltage input. 

The APFC converter needs tuning for higher PF readings.

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Deepcool trusted Seasonic's trendy Focus Gold platform for the PQ-M series, which is starting to show its age, especially now that the ATX12V v3.0 is out. There is no 12+4 pin connector, so that the PQ850M won't support the upcoming GPUs, and adapters are not recommended since they don't pass through the sense signals of the 12VHPWR connectors, which will unlock the full potential of the next-generation graphics cards. 

The amount of PCIe connectors might trouble some of you. Still, nowadays, there is no need to install more than two graphics cards in gaming systems, so three PCIe connectors on dedicated cables will cover 99% of usage scenarios. There are also two EPS connectors available for power-hungry CPU and mainboard combinations. 

The PQ850M's overall performance is not far from most competing offerings, and it is among the quietest Gold PSUs with 850W capacity. This is not enough to dethrone the mighty Corsair RM850x, which sets the performance bar high in this category. It would help if you also kept in mind that we are in a transition phase right now, where every PSU with more than 450W capacity has to be equipped with a 12VHPWR connector to support the upcoming GPUs along with the rest changes that the ATX12V v3.0 spec requires. We expect most manufacturers to update the existing PSU models soon, so you better wait if you are not in a dead hurry for a new power supply.

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Disclaimer: Aris Mpitziopoulos is Tom's Hardware's PSU reviewer. He is also the Chief Testing Engineer of Cybenetics and developed the Cybenetics certification methodologies apart from his role on Tom's Hardware. Neither Tom's Hardware nor its parent company, Future PLC, are financially involved with Cybenetics. Aris does not perform the actual certifications for Cybenetics.

We recently took a look at DeepCool’s CG560, an $85 mid-tower case with three 120mm RGB fans and a 140mm exhaust and came away mostly impressed. But it’s a new year and the company is back with two new cases: the CK560 and the CK500, which are also budget-friendly, air-flow-focused mid-towers, along with a couple new peripherals--a first for the company mostly known for cases and cooling.

The new CK560 still features that reticle pattern on the front panel (although it’s metal this time, and magnetically attached). It also has three RGB 120mm fans and one 140mm exhaust (sans molex connector this time). 

Aside from generous and well-spaced top ports, there isn’t a ton of differences between this version and the CG model. DeepCool did, however, push the front panel out a bit more to allow for more breathing room for your hardware and add in an anti-sag GPU support bracket.

Next up is the CK500, which has a more minimal design, as it lacks that reticle-like design up front, but is still, seemingly pushed out enough to allow for better airflow than most sealed front panel cases.

The CK560 retails for $99 in black, or $104 in white, while its sibling case, the CK500 goes for $80.

MG510 Mouse and KG722 65 Mechanical Keyboard

Don’t go anywhere, because DeepCool also has some new peripherals on offer: The MG510 gaming mouse and the KG722 65 percent mechanical keyboard.

Sliding in first with its Ultra-Glide mouse feet is the MG510, which features a PixArt PAW3370 optical sensor, 19,000 DPI, 2.4 GHz wireless, 36 hours of battery life, six programmable buttons, aRGB lighting and OMRON Micro Switches.

Along with its high DPI, the PixArt sensor allows for an IPS of 400 and a 1000hz polling rate, all for $80.

Clicking and clacking in is DeepCool’s new 65 percent mechanical plank, the KG772. DeepCool’s new wired keyboard is fitted with Gateron Red mechanical switches, which will provide you with a linear keystroke at an operating force of 45g. Covering Gateron’s switches are laser-engraved ABS, RGB shine-through keycaps that are side-printed, giving your keyboard an understated look. The keycaps are attractive, but very smooth. Likewise, the board feels solid, but is mostly featureless on its sides, as this is a wired model without the need for any buttons or toggle switches.

While the keycaps are discrete, the KG772 has per-key RGB, which can be addressed with DeepCool’s own software. We just wish it had flip-up feet underneath for extra ergonomic versatility.

Just like the MG510, the new KG722 retails for $80, and both should be on sale right now.

Deepcool's new single 140mm air cooler, the AS500, resembles half of the company's Assassin III large tower cooler. If the nickel-plated cooling heatpipes and black accent trim on the AS500 attributes to its proper, tuxedo-like appearance, consider the lighted accent top plate the dynamic aRGB bow tie.

The Deepcool AS500 will be available in UK (£54.99) and European markets around October 20, 2020, arriving sometime in November in the US, for $59.99. With simple installation and support for most current Intel and AMD CPUs (sorry, no Threadripper), the AS500 shines as an intermediate option between budget and large air cooling options.

Deepcool AS500 Specifications

Deepcool ships the AS500 with an assortment of nickel-plated mounting hardware and a universal backplate which provides mounting compatibility with most current Intel and AMD CPU sockets, although AMD Threadripper does not make the list. Spring clips are provided for those wishing to run an additional fan on their AS500 for push/pull operation, while a syringe of thermal paste and an aRGB control module round out the accessory package.

Deepcool covers the AS500 with a 3-year warranty.

The AS500 features a set of five nickelplated heatpipes which converge at the base of the cooler between the mounting plate and a milled baseplate. Tension screws are fixed within the mounting plate bracket, providing simplified installation.

Deepcool provides a 140mm TF140S PWM fan rated at 70.8 CFM at 1200 RPM. The trailing edges of the fan blades are trimmed with fins for directional airflow through the fan chassis and into the cooling tower. The fan itself is not aRGB capable, but it does come equipped with rubberized fan mounts for minimizing vibration noise.

Deepcool makes use of offset heatpipes within the aluminum cooling tower to increase airflow around each thermal conduit. The top bezel of the AS500 is trimmed with an aRGB opaque diffuser and topped with a black top piece. Cabling for the aRGB lighting module runs down a vertical channel cut into the cooling tower to prevent interference of the cabling with the 140mm fan.

In checking the milling of the baseplate, we can see just a slight bit of convexity, which meets the CPU integrated heat spreader (IHS). This should not be an issue for most installations but could amplify a problem with improper installation or seating of the cooler atop the CPU.

Our installation resulted in an even distribution of thermal compound across the entire IHS face of our Core i9-10850K CPU. This is precisely what we like to see following a cooler install.

Since the Deepcool AS500 is a single tower cooler, it is centered over the CPU and allows for wider compatibility with memory DIMMs which might feature tall heat spreaders, like our Corsair Dominator DDR4-3600 modules, but some installations might still bump into this problem, depending on your configuration and motherboard.

The installation of the AS500 is quite simple, making use of mounting crossbars and the fixed tension springs on the mounting plate which try to automatically align the cooler correctly over the mounting brace.

While not completely immune to memory interference, the AS500 does a better job than most 140mm coolers. This is nearly always an issue with CPU air cooling solutions featuring large fans, so make sure to do a bit of research and shop accordingly.

Test System Setup

For our CPU cooling tests, we use the same hardware, overclock and configuration for each and every test to minimize environment variables in testing. This allows for all results across all coolers tested on the platform to be viable as side-by-side examination for direct compare/contrast.

Data comparisons are based on data collected from testing performed on our new Intel i9-10850K system, including re-visiting many previously covered products which were originally covered on the prior testing platform which pivoted around an i7-5930k (4.20ghz @1.20v).

All data reported for this article has been collected on the current Intel i9-10850K platform and will be maintained as like-for-like evaluation of ongoing cooling coverage.

Prime95 v29.4b8 (no AVX) is used for two-hour intervals, one managing fans at 50% PWM and the other at 100% PWM with RPM measurements being taken every 3 seconds and averaged across the duration of each 2-hour capture. Omitting AVX instruction sets allows for accurate, 100% loads at chosen clock speeds while allowing AVX instructions would provide higher, albeit, unrealistic synthetic CPU loads and excessive heat production, less indicative of real-world use.

This also allows for a greater range of CPU coolers to be tested and compared without the need to configure the system differently for smaller coolers, which may not handle the excessive thermal loads being generated during testing, while larger coolers might be better equipped to manage heat output produced by the Core i9-10850K .

While the test platform is quite capable of a 10-core overclock at 5.0Ghz and 1.265v, we were seeing 360mm AIOs struggle to keep core temperatures in check at lower fan speeds, providing insight that the enthusiast-grade i9’s need excellent cooling if the goal is substantial overclocking.

HWInfo64 is used for real-time core temperature readout, thermal throttling alerts, motherboard power consumption, CPU speed and logging of data, while a CrystalFontz CFA-633-TMI-KU is used to monitor and later average both ambient room (2 probes) and motherboard voltage regulator heatsink (2 probes).

Comparison Coolers

Looking at our thermal load temperature comparison, right away we see some strong similarities to the Deepcool AS500 with its larger sibling, the Assassin III. The Cooler Master Master Air MA410M and be quiet! Shadow Rock 3 are each similar in price as well as overall performance, making them good comparisons for this testing group.

As expected, the larger 140mm fans of the Deepcool products rotate a bit more slowly than the faster pair of 120mm fans used by the MA410M and the 135mm fan used by the Shadow Rock 3.

Large, 140mm fans like those on the Deepcool AS500 and Assassin III benefit from slower rotational speed, but even the 135mm fan of the be quiet! Shadow Rock 3 makes an impact with low decibel levels. However, even a pair of 120mm fans on the Cooler Master MA410M are reasonably good at both full and half speeds.

The Deepcool AS500 benefits from doing very well in our thermal load tests while remaining very quiet in the process. Close comparisons between thermal load temperatures or noise levels can result in some interesting results within this chart.

Conclusion

The Deepcool AS500 is priced at $59.99 (£54.99) at launch, making it a very good performance value compared with others in the testing group. At that price, this cooler looks to be an early system builder favorite.

Shots from our FLIR ONE Pro thermal camera show only slight differences between 50% and 100% fan speed operation under load. The AS500’s 140mm fan and cooler tower provide efficient cooling at nearly inaudible noise levels.

Deepcool impresses with its new 140mm tower cooler, the AS500. At first glance, it appears to be simply half of the Assassin III dual tower cooler, but we hesitate to even make this comparison as the AS500 performs within just a few degrees of its larger sibling in our tests while doing so at a substantial price discount.

It should be noted that higher core counts and stronger overclocks will ultimately begin to push the AS500 beyond its limit and into the realm of large, dual-tower air cooling, 360 AIOs or custom water cooling. However, for the vast majority of system builders, the AS500 provides great results as a less-costly alternative to higher-priced extreme air and AIO coolers.

The DeepCool GamerStorm DQ-M V2L 850W uses a new platform from Channel Well Technology (CWT), which has the codename GPX. In essence, GPX platform is a downgraded GPU design (which we have seen in the Bitfenix Whisper line and the older DeepCool GamerStorm models). Its goal is to offer high enough performance at a lower price, meaning that it doesn't use the same high-quality parts that the GPU design utilizes. With higher quality FETs and a stronger boost diode, the GamerStorm DQ-M V2L 850W could be included in our Best Power Supplies article, but performance alone isn't enough. 

The design needs to be reliable enough to last for years to come, even under tough operating conditions, and it should also retain its good performance over time, because even PSUs age. DeepCool's new power supply line looks interesting, but it doesn't pose any threat to the highly popular Corsair RMx and Seasonic Focus Plus Gold lines.

DeepCool decided to expand its GamerStorm brand with a new PSU line, with the codename DQ-M V2L. This line consists of three members with capacities ranging from 650W to 850W. Compared to the original DQ-M line, which is based on the top-notch CWT GPU platform, the DQ-M V2L uses the inferior CWT GPX platform with some modifications, which allow for higher performance and increased reliability. The latter is depicted on the ten-year warranty that supports these new PSU models. Finally, all DQ-M V2L units feature a single +12V rail design, contrary to the DQ-M members, which had several +12V rails.

The GamerStorm DQ-M V2L 850W is the flagship of the line, featuring enough power (and connectors) to support a potent gaming system. It is a fully modular power supply with regular dimensions. A few years ago a 850W PSU with 160mm depth would be considered compact, but nowadays, we find similar capacity PSUs with only 140mm depth. With a quick look at the specifications, we wonder why DeepCool didn't use a larger fan (e.g., 140mm instead of 120mm), since the chassis is large enough to accommodate it. A larger fan can provide the same airflow with a smaller one, but at lower speeds so its noise output will be lower.

Specifications

Power Specifications

Cables & Connectors

The ATX cable could be a bit longer, reaching 600mm, to ensure compatibility with full tower chassis. The length of the EPS connectors is satisfactory, but 50mm more wouldn't hurt. 

It isn't common to see six 4-pin Molex connectors, even in 850W PSUs. DeepCool thought that users would like to have a large number of these connectors to power peripheral devices that need more juice than what SATA connectors can provide. It is nice also to see a large enough distance between peripheral connectors. 

Component Analysis

We strongly encourage you to have a look at our PSUs 101 article, which provides valuable information about PSUs and their operation, allowing you to better understand the components we're about to discuss.

The GPX platform is a downgraded version of the GPU design, so we didn't expect to find top-notch quality parts and, in some areas, e.g., in the APFC converter, indeed this is the case. Compared to other 850W units with similar specs, this one uses a notably weaker boost diode in the APFC circuit, and on top of that, the quality of the FETs that CWT used is not high. For example, the Silan Microelectronics FETs are also used in the much more affordable Corsair CX450 (the one made by CWT). It would be ideal if CWT used Infineon or On Semiconductor FETs, but this would affect the production cost.

The good thing is that a full-bridge topology is used, while the GPU platform uses half-bridge. Briefly, a full-bridge topology can deliver more power, with the proper components, and it has lower energy losses than a half-bridge configuration. Nonetheless, in radiated EMI, the half-bridge converter has the edge over the full-bridge converter.

The transient filter includes all necessary components to restrict conducted EMI effectively. 

The single bridge rectifier is bolted on the primary heatsink. The DQ850-M, based on the GPU platform, also uses a single bridge rectifier with similar specifications. 

In the APFC converter, we find two Great Power FETs, with model number GP28S50 and a single boost diode, which might be provided by a good manufacturer, but it isn't so strong. Usually, we find 8A boost diodes in higher-end 850W units. The DQ-850M uses an Infineon IDH08G65C5 (650V, 8A @ 145°C), for example. 

It is nice to see a full-bridge topology, but we cannot say the same for the primary switching FETs. To keep the cost down, CWT used more affordable FETs compared to the ones that the DQ850-M has in the same stage. To speak with numbers, two Fairchild FCPF125N65S3 FETs (DQ850-M) cost $5.62 while four Silan Microelectronics SVF20N50F go for $3.26. Saving $2.36 in the primary FETs alone is a huge deal! 

The +12V rail uses six InPower Semiconductor FETs, while the pair of DC-DC converters that handle the minor rails utilize four Sync Power FETs. 

The majority of electrolytic filtering caps are by Chemi-Con. Besides the high-end KY and KYA lines, we also find four lower-end KZE and KMG caps, along with a single Nichicon VZ, with only 1,000h lifetime. CWT also used a large number of polymer caps, which are made by FPCAP.

The standby circuit is controlled by a Power Integrations TNY287PG IC. 

The supervisor IC is provided by Sitronix and supports all necessary protection features but OTP (Over Temperature Protection), which is implemented through another circuit. 

Several polymer caps are installed at the face of the modular board, along with several bus-bars that transfer power to the modular sockets. 

The soldering quality is very good. CWT has solid production lines.

Most manufacturers have turned to Hong Hua because it offers high performance per buck products with satisfactory quality. This specific fan uses a rifle bearing, so under normal conditions, it will last for quite long. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

To learn more about our PSU tests and methodology, please check out How We Test Power Supply Units. 

Primary Rails And 5VSB Load Regulation

The following charts show the main rails' voltage values recorded between a range of 40W up to the PSU's maximum specified load, along with the deviation (in percent). Tight regulation is an important consideration every time we review a power supply because it facilitates constant voltage levels despite varying loads. Tight load regulation also, among other factors, improves the system’s stability, especially under overclocked conditions and, at the same time, it applies less stress to the DC-DC converters that many system components utilize.

Load regulation is tight on all rails but 5VSB, where it doesn't play a significant role. 

Hold-Up Time

Put simply; hold-up time is the amount of time that the system can continue to run without shutting down or rebooting during a power interruption.

The hold-up time is much longer than what the ATX spec requires (17ms), and the power ok signal is accurate. 

Inrush Current

Inrush current, or switch-on surge, refers to the maximum, instantaneous input current drawn by an electrical device when it is first turned on. A large enough inrush current can cause circuit breakers and fuses to trip. It can also damage switches, relays, and bridge rectifiers. As a result, the lower the inrush current of a PSU right as it is turned on, the better.

The inrush current is low with 115V and quite high, but still not dangerously high, with 230V. 

10-110% Load Tests

These tests reveal the PSU's load regulation and efficiency levels under high ambient temperatures. They also show how the fan speed profile behaves under increased operating temperatures.

The efficiency levels get a big hit under high operating temperatures, showing that the parts that CWT used are not highly tolerant to harsh operating conditions. The cooling fan also has to work at its maximum speed from test #9, to cope with the thermal load and keep as cool as possible the low-quality FETs.

20-80W Load Tests

In the following tests, we measure the PSU's efficiency at loads significantly lower than 10% of its maximum capacity (the lowest load the 80 PLUS standard measures). This is important for representing when a PC is idle with power-saving features turned on.

With 20W load, efficiency doesn't exceed 70%. In all other tests, it exceeds 80%, though. 

2% or 10W Load Test

Intel plans on raising the ante at efficiency levels under ultra-light loads. So from July 2020, the ATX spec will require 70% and higher efficiency with 115V input. The applied load is only 10W for PSUs with 500W and lower capacities, while for stronger units we dial 2% of their max-rated-capacity.

With 2% load, the 70% efficiency mark isn't reached, so this PSU cannot meet the newest ATX spec requirements. 

Efficiency

Next, we plotted a chart showing the PSU’s efficiency at low loads, and loads from 10 to 110% of its maximum-rated capacity. The higher a PSU’s efficiency, the less energy goes wasted, leading to a reduced carbon footprint and lower electricity bills.

The efficiency levels are at satisfactory levels, given the parts that this platform uses. 

5VSB Efficiency

The 5VSB rail is highly efficient! 

Power Consumption In Idle And Standby

The vampire power levels are low, with both voltage inputs. 

Fan RPM, Delta Temperature, And Output Noise

All results are obtained between an ambient temperature of 37 to 47 degrees Celsius (98.6 to 116.6 degrees Fahrenheit).

Up to 500W, the fan spins at below 1000 RPM, but at higher loads, the fan speed profile gets wild. The low-quality FETs don't allow for a more relaxed speed profile, as it seems. 

The following results were obtained at 30 to 32 degrees Celsius (86 to 89.6 degrees Fahrenheit) ambient temperature.       

There is no passive operation, but this is not a problem since the fan isn't noisy, with loads up to around 550W. With higher loads, though, the fan quickly increases its speed, and the same goes for the output noise, which exceeds 40 dBA with >610W loads. In some cases, noise can go over 45 dBA, so if you plan to push this PSU hard, you should be prepared for noisy operation. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Protection Features

Check out our PSUs 101 article to learn more about PSU protection features.

The over current protection at +12V along with over power protection is configured correctly. This doesn't apply, though, to the minor rails which are close to 145%, compared to their nominal levels. There is no need for such high power levels on these rails and at some point, PSU manufacturers should realize it. The other protection features have normal operation. 

DC Power Sequencing

According to Intel’s most recent Power Supply Design Guide (revision 1.4), the +12V and 5V outputs must be equal to or greater than the 3.3V rail at all times. Unfortunately, Intel doesn't mention why it is so important to always keep the 3.3V rail's voltage lower than the levels of the other two outputs.

The 3.3V rail is lower than the other two in all tests we conducted, so there are no problems here. 

Cross Load Tests

To generate the following charts, we set our loaders to auto mode through custom-made software before trying more than 25,000 possible load combinations with the +12V, 5V, and 3.3V rails. The deviations in each of the charts below are calculated by taking the nominal values of the rails (12V, 5V, and 3.3V) as point zero. The ambient temperature during testing was between 30 to 32 degrees Celsius (86 to 89.6 degrees Fahrenheit).

Load Regulation Charts

Efficiency Chart

Ripple Charts

The lower the power supply's ripple, the more stable the system will be and less stress will also be applied to its components.

Infrared Images

We apply a half-load for 10 minutes with the PSU's top cover and cooling fan removed before taking photos with a modified FLIR E4 camera able to deliver an IR resolution of 320x240 (76,800 pixels).

The hottest areas are near the bridge rectifier and the voltage regulators that handle the minor rails. The temperatures are not high, but CWT was forced to use an aggressive fan speed profile to keep as cool as possible the internals, to ensure that the low-quality FETs will outlive the provided warranty. The latter is super-long for the low-grade FETs under mid to severe operating conditions. Only time will reveal more on this platform's reliability. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Advanced Transient Response Tests

For details about our transient response testing, please click here.

In the real world, power supplies are always working with loads that change. It's of immense importance, then, for the PSU to keep its rails within the ATX specification's defined ranges. The smaller the deviations, the more stable your PC will be with less stress applied to its components. 

We should note that the ATX spec requires capacitive loading during the transient rests, but in our methodology, we also choose to apply a worst case scenario with no additional capacitance on the rails. 

Advanced Transient Response at 20% – 20ms

Advanced Transient Response at 20% – 10ms

Advanced Transient Response at 20% – 1ms

Advanced Transient Response at 50% – 20ms

Advanced Transient Response at 50% – 10ms

Advanced Transient Response at 50% – 1ms

Only 5V and 5VSB perform well in these tests. The +12V is away from 1% deviation, which is considered good performance while the deviation at 3.3V exceeds 4%. 

Turn-On Transient Tests

In the next set of tests, we measure the PSU's response in simpler transient load scenarios—during its power-on phase. Ideally, we don't want to see any voltage overshoots or spikes since those put a lot of stress on the DC-DC converters of installed components.

We only noticed a small spike at 5VSB and two small steps during the last test. All in all, good enough performance here. 

Power Supply Timing Tests

There are several signals generated by the power supply, which need to be within specified, by the ATX spec, ranges. If they are not, there can be compatibility issues with other system parts, especially mainboards. From year 2020, the PSU's Power-on time (T1) has to be lower than 150ms and the PWR_OK delay (T3) from 100 to 150ms, to be compatible with the Alternative Sleep Mode.

The PWR_OK delay is out of the 100-150ms region, so the PSU does not support the alternative sleep mode (ASM). No mainboards are supporting this mode yet, but given the ten-year warranty, you will keep this PSU for quite long, so it would be ideal for supporting ASM.

Ripple Measurements

Ripple represent the AC fluctuations (periodic) and noise (random) found in the PSU's DC rails. This phenomenon significantly decreases the capacitors' lifespan because it causes them to run hotter. A 10-degree Celsius increase can cut into a cap's useful life by 50%. Ripple also plays an important role in overall system stability, especially when overclocking is involved.

The ripple limits, according to the ATX specification, are 120mV (+12V) and 50mV (5V, 3.3V, and 5VSB).

The ripple suppression is good on all rails, even at 5V, which looks high in the charts compared to other, similar spec, PSUs. From the moment the minor rails stay below 20mV, at full load and higher than 45 degrees Celsius ambient, we cannot complain about their performance. 

Ripple At Full Load

Ripple At 110% Load

Ripple At Cross-Load 1

Ripple At Cross-Load 2

EMC Pre-Compliance Testing – Average & Quasi-Peak EMI Detector Results

Electromagnetic Compatibility (EMC) is the ability of a device to operate properly in its environment without disrupting the proper operation of other nearby devices.

Electromagnetic Interference (EMI) stands for the electromagnetic energy a device emits, and it can cause problems in other nearby devices if too high. For example, it can be the cause of increased static noise in your headphones or/and speakers.

With the Average detector several, sixteen to be exact, spurs exceed the limits in the 342 - 808 kHz range. Using the QP detector, which is more accurate, everything is ok, although some emissions are dead close to the limits. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Performance Rating

The overall performance is very good, taking the lead from the XPG Core Reactor 850, which has higher build quality, though. The RM850x and Seasonic Focus Plus Gold with similar capacity perform better than DeepCool's offering. 

Noise Rating

The graph below depicts the cooling fan's average noise over the PSU's operating range, with an ambient temperature between 30 to 32 degrees Celsius (86 to 89.6 degrees Fahrenheit).

This is a noisy PSU. The DQ850-M, which uses CWT's GPU platform (GPU850V-G), has a much more relaxed fan speed profile since it uses higher quality parts, which are more tolerant to high operating temperatures. 

Efficiency Rating

The following graph shows the PSU's average efficiency throughout its operating range with an ambient temperature close to 30 degrees Celsius.

The overall efficiency is a bit higher than the DQ850-M that uses a half-bridge topology. This is an efficient platform, but given the full-bridge topology, we expected even better results here. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

DeepCool utilized a mainstream platform from CWT, with code name GPX, and with some modifications, it managed to get good performance out of it along with a long hold-up time, since they also changed the bulk capacitor. This change costs a lot, since bulk caps are among the most expensive parts in a power supply, along with the cooling fan and the switching MOS(FETs). 

Speaking of FETs, we were not pleased, at all, to see lower grade FETs made by off-brand Chinese manufacturers in this unit. CWT might use a superior full-bridge topology in this platform, instead of the half-bridge that the higher-grade GPU design uses. Still, it only did so to lower the production cost since this topology allowed for lower and much more affordable FETs. It is strange to see such an extended warranty period, reaching ten-years, in a product that doesn't use quality FETs. The aggressive fan speed profile might boost reliability, but it won't help in the long run if you push hard the PSU on a daily basis. 

With higher quality FETs and a stronger boost diode in the APFC converter, this power supply would be equally good, if not better, than its competitors. This is not the case, though, and despite the ten-year warranty that DeepCool provides, we are not sure if it will manage to outlive it, especially if you are not gentle with it. 

On top of that, the fan speed profile is aggressive, so under high loads and increased operating temperatures, the PSU will make its presence felt. This product could be so much better if it used the proper parts. As it is, the majority of competing offerings, including the Corsair RMx, Seasonic Focus, and XPG Core Reactor, are better choices. 

MORE: Best Power Supplies

MORE: How We Test Power Supplies

MORE: All Power Supply Content

Disclaimer: Aris Mpitziopoulos is Tom's Hardware's PSU reviewer. He is also the Chief Testing Engineer of Cybenetics and developed the Cybenetics certification methodologies apart from his role on Tom's Hardware. Neither Tom's Hardware nor its parent company, Future PLC, are financially involved with Cybenetics. Aris does not perform the actual certifications for Cybenetics.

If you're in the market for a big all-in-one liquid cooler, perhaps the Deepcool Castle 280EX is worth keeping an eye on. Being a 280 mm unit, this one comes with significantly more radiator area than a 240 mm unit, which helps increase cooling performance. 

The Castle 280EX comes with a 3-phase motor that is redesigned to offer an increased number of cold plate fins for better cooling performance across its surface underneath. Heat is removed from the radiator by two Deepcool TF140 S fans.

Fan noise is quoted to remain under 39.8 dBA at the full speed of 1600 RPM, though chances are you won't ever need to run them that fast unless you're overclocking the super-hot i9-10900K to its limits.

The all-in-one liquid cooler comes with Deepcool's own anti-leak system, which essentially adds a rubber bubble to the radiator that can expand and retract, equalizing the unit's internal pressure to the environment.

Included with the DeepCool Castle 280EX is an RGB controller for the pump/cold plate combo unit, though you can also choose to control the lighting through your motherboard. The top of the pump also turns, so you can orient it to match your system.

Pricing is set at $150 with availability slated for this month.

DeepCool's Gammaxx GT CPU cooler came out about two years ago, and at the time it featured RGB lighting that could handle a single color at a time. Nowadays, that isn't enough, so the vendor has updated the cooler to include addressable RGB (ARGB) and gave it a small make-over. The updated component is dubbed the DeepCool Gammaxx GT A-RGB.

The CPU cooler is based on a classic single-tower design, which uses four 6mm direct-contact heatpipes made from copper, which lead to an aluminum fin stack. Attached to this is a 120mm fan that can spin at speeds between 500 and 1,650 RPM. At max speed it can push up to 64.5CFM while making 27.8dBA of noise, and it features PWM fan control. 

The ARGB is featured in both the fan as well as the top of the cooler, where the DeepCool logo can be lit up in various colors of choice. An adapter is included to connect the whole lot.

All that leads to a cooler measuring 129mm wide and 77mm deep, so it should have perfectly good RAM clearance. It's also 157mm tall, so it'll also fit in most PC cases, except those that are unusually small.

We don't have information on pricing yet, but the Gammaxx GT cooler sells for around $40 to $60, depending on the variant you choose, so we reckon the Gammaxx GT A-RGB will be around that price range once inventories of the old units clear.

Deepcool recently announced its Gamer Storm Castle 360EX White CPU cooler, which is a triple-fan 360mm all-in-one liquid cooling system.

On the face of it, this liquid cooler doesn’t appear to contain anything all too special. It looks like a pretty straight-forward 360mm all-in-one unit, albeit with some nifty white styling and fancy lighting on the CPU block. However, Deepcool has integrated a leak-prevention technology that it calls “Anti-Leak Tech Inside.” Anyone else getting an “Intel Inside” marketing vibe here?

Deepcool's leak-prevention tech works by including a sort of rubber baggy that can expand and shrink inside a space at the end of the radiator. One side holds coolant, and the other side is exposed to the environment, allowing the internal pressure of the unit to stay somewhat close to the environmental pressure – essentially, the leak prevention works by offering a form of pressure relief.

We’ve discussed this feature internally here at Tom’s Hardware, and we’re not convinced it offers any added value. The rubber tubing on most AIO’s these days has enough expansion capacity to deal with any of these changes, and even something as simple as an air bubble that gets bled to a small reservoir is adequate to deal with any of the pressure changes in these sealed units. This doesn’t mean that it’s a bad feature per se, but rather that we don’t believe that these units are any more leak-resistant than most properly-built competitors -- so be careful before buying into the marketing materials too deeply.

The unit comes with three TF120-S fans that come with the same color frame as the radiator. They have a second layer at the end of the blades, and come with noise-damping anti-vibration pads at the corners of the frame. They’ll spin at speeds between 500 and 1800 RPM, and will make up to 32.1 dBA of noise while pushing up to 64.4CMF per fan. Fortunately you can control them through PWM to keep those noise levels down during light workloads.

The unit’s pump features a 3-phase motor that consumes up to 2.4W. It isn’t PWM controllable with just a 3-pin connector, but also makes only 7.8 dBA of noise at its standard speed of 2550 RPM, so it shouldn’t be too much of a bother anyway. The lighting on the pump is addressable RGB, and it has replaceable internal logos, so you can replace this with the included ying-yang template or create your own.

The Deepcool Gamer Storm Castle 360EX White is compatible with all modern-day sockets, including AMD’s TR4 socket.

Deepcool assigns the cooler a $169.99 MSRP, and they’re currently available for purchase.

Features and Specifications

DeepCool's Gamer Storm Castle 240EX AIO liquid cooler sports a handsome, mirror-faced design with aRGB backlighting and a design meant to preventatively purge excess pressure within the cooler itself in order to stave off coolant leaks. Cooled by a pair of non-aRGB 120mm PWM fans, the Castle 240EX provides notable cooling performance and colorful aesthetics.

Priced at $120, the DeepCool Gamer Storm Castle 240EX aligns with other premium 240mm AIO options which tend to focus on both thermal performance and aRGB/RGB presentation. When compared side-by-side among these, it provides relatively average thermal results, simple installation and limited aRGB controls at a moderately higher price point.

DeepCool Gamer Storm Castle 240EX Specifications