A Chinese device maker is teasing a new mini PC inside a chassis that bears more than a passing resemblance to the Nintendo Entertainment System (NES). Acemagic’s blog and social media share images and a handful of specs for this AMD Ryzen AI 9 HX 370 machine, which it has dubbed the Retro X5. The maker characterizes this mini PC as “an all-in-one solution created to simplify classic gaming,” which makes us wonder if it will come with Batocera Linux, SteamOS, or another non-Windows OS/UI.

We don’t know a lot about the hardware inside the Acemagic Retro X5 as yet. The Ryzen AI 9 HX 370 inside is a capable processor, though. It features a 4x Zen 5 performance cores and 8x Zen 5c efficiency cores for a total of 12 cores and 24 threads, running at up to 5.1GHz. Onboard is a decent iGPU in the shape of the Radeon 890M (RDNA 3+). This packs in 16 CUs and runs at up to 2.9 GHz. There’s an XDNA 2 NPU with 50 TOPS, too.

If we are thinking about PC gaming with this processor, it should easily overshadow the Steam Deck. However, you will be targeting higher resolutions on a desktop. Also, as a mini PC, the HX 370 should run without the thermal constraints typically experienced in handheld or laptop designs. Thus, the Retro X5 might push this processor to the upper part of the processor’s configurable 15-54W TDP range.

From the few images of the new mini PC we can see a USB-C, 2x Type-A, a 3.5mm audio jack, and a power button up front. We also know that the “Retro X5 supports expandable memory and storage, multiple display outputs, and high-speed connectivity,” as it says so on the blog. Hopefully, that means at least: twin SODIMM slots, twin M.2 SSD slots, HDMI, USB-C monitor connectivity, USB4 or USB5, OCuLink, and Ethernet, multiple extra ports along the back, and built-in Wi-Fi / BT.

RetroPlay Box

It looks like Acemagic will try and differentiate its latest mini PC with some kind of launcher, dubbed RetroPlay Box, which you can see represented by swirling tiles in the Tweet embedded above. It could be an app that is pre-installed on Windows 11 – the least adventurous option.

There’s also the chance that Acemagic might go with a Linux-based OS to host its RetroPlay Box. It may save money on a Windows license by choosing SteamOS or Batocera Linux, for example. We shall have to wait and see.

Surfing the retro-wave

Acemagic is riding on the coattails of Ayaneo (and Nintendo) with its “inspired by iconic gaming consoles” design here. But, perhaps its timing, surfing the wave of retro-enthusiasm precipitated by Windows 11 disillusion, the rise of SteamOS, and yet another PC AAA gaming hardware cost crunch, will help it achieve better success.

The Acemagic Retro X5 is “coming soon,” according to the firm’s social media postings.

Tom’s Hardware has tested dozens of processors from AMD and Intel to find the best CPU for gaming. Our list of 2026 CPU gaming benchmarks currently comprises 17 of the most demanding titles available on the market, which we run each gaming processor through to see the chips that come out on top. We select our picks based on the data in our CPU benchmark hierarchy, so all of the CPUs below are backed by hundreds of hours of real-world, hands-on testing where we gather extensive data on how a CPU performs and behaves while gaming. If you want a broader look at the CPU market, our AMD vs. Intel article shows you where the current CPU duopoly stands, while our CPU buying guide can help you narrow down the best processor for you.

Now in the back half of the year, we don’t expect major new releases from AMD or Intel. AMD has been on a tear with refreshes, particularly among X3D CPUs. We recently reviewed the Ryzen 9 9950X3D2 Dual Edition, AMD’s first CPU with 3D V-Cache on both CCDs, and it’s the most powerful chip from Team Red currently available. It’s earned a spot on this list, though the (much cheaper) Ryzen 9 9950X3D isn’t far behind in overall performance.

We’ve also seen the Ryzen 7 9850X3D, which is technically the fastest gaming processor on the market, as you can see in our Ryzen 7 9850X3D review. However, we’ve kept the base Ryzen 7 9800X3D as our top recommendation for the best CPU for gaming due to its price. It’s only marginally behind the refreshed model (about 3% on average), and much cheaper. For most gamers, the Ryzen 7 9800X3D makes more sense.

Intel has seen a recent boost in gaming performance with Arrow Lake Refresh, and both the Core Ultra 5 250K Plus and Core Ultra 7 270K Plus have earned spots on our list. AMD dominates in gaming at the moment, however. Our sights are set on Intel’s next-gen Nova Lake chips for a big gaming boost from Team Blue, as well as the rumored ‘Raptor Lake Next’ lineup that’s supposedly arriving early next year.

For the rest of the year, we have the Ryzen 7 7700X3D and Ryzen 7 5800X3D 10th Anniversary Edition to look forward to, which are arriving in July and June, respectively. We don’t anticipate they’ll make a major impact on our rankings here, but we plan on reviewing both CPUs as soon as they’re available.

In addition to the fastest CPUs from AMD and Intel, we’ve included a few DDR4 options on this list. The price of DRAM and NAND flash has made building even a budget PC prohibitively expensive, so DDR4 platforms are a great way to save money. Vendors are signaling a shift back toward DDR4 platforms at the moment, so we may reconsider some older CPUs for our rankings as pricing and availability allows.

Here are the gaming CPUs we recommend buying. We have a shortlist of the top options and some alternatives below, but you can click the ‘More’ links to read our thoughts about a particular CPU and where it stands in the current market.

Best CPU for Gaming in 2026 at a glance (more info below):

The list below is for the best CPUs for gaming, while the best budget CPUs can help you find a cheap chip. Processors benefit from the best thermal paste, so check out our guide if you're shopping for a new processor. But if you're after the best CPU for gaming, you're in the right place.

Best CPU for Gaming Benchmarks

We rank all the Intel and AMD processors based on our in-depth CPU benchmarks hierarchy. You can see some of those numbers in the charts above, including CPU overclock performance results (marked as PBO for AMD processors). We're currently retesting all of these processors with the Nvidia RTX 5090, but only the first four slides have that testing. The remainder are historical testing results with the RTX 4090, which we'll remove once we have fully retested all of the gaming CPUs with the RTX 5090 for our benchmarks. This group of results comprises only the chips that have passed through our newest test suite. Additionally, the tables in our CPU benchmark hierarchy include rankings based on past CPU benchmarks and breakdowns of single- and multi-threaded performance in productivity applications across a broad spate of processors. Finally, be aware that the pricing in the charts above can fluctuate.

Quick Shopping Tips

When choosing the best CPU for gaming in 2026, consider the following:

• You can't lose with AMD or Intel: As noted in our AMD vs. Intel feature, AMD tends to make the best all-around CPU for gaming for mainstream PCs lately, but both offer compelling performance options at any given price point.
• Eight cores is sufficient for gaming: If you’re looking at a pricey flagship, you’re likely wasting some money if gaming is your primary focus. You can game on as little as a quad-core CPU, but performance scaling really falls off past eight cores.
• Budget platform costs: You never want to pair a strong CPU with a weak GPU, RAM, and storage. Right now, it’s especially important to consider platform costs, however. DDR5 prices are peaking, and you’ll need to factor in the cost of DDR5 and a new motherboard if you’re coming from an older socket like AM4.
• Overclocking isn’t for everyone, but if you follow our How to Overclock a CPU guide, you can scrape out extra performance gains.

Best CPU for Gaming 2026 - $300 to $400

The Ryzen 7 9800X3D technically isn’t the fastest gaming chip on the market any more. That title goes to the new Ryzen 7 9850X3D, though the victory is marginal. As you can read in our Ryzen 7 9850X3D review, AMD’s latest X3D offering pushes ahead by 3.3% on average. Despite a minor uplift, we’re still recommending the Ryzen 7 9800X3D. With prices as they currently are, the Ryzen 7 9850X3D is only 3.3% faster despite costing around 6% more than the Ryzen 7 9800X3D.

This chip really has no peer in the market outside of the Ryzen 7 9850X3D — the Ryzen 7 9800X3D delivers outstanding gaming performance, beating Intel's fastest gaming chip, the $469 Core i9-14900K, by 30% in our test suite. The 9800X3D is also almost unbelievably 35% faster than the current-gen Intel flagship, the $560 Core Ultra 9 285K. The stock Ryzen 7 9800X3D's 1% low frame rates (a good smoothness indicator) also deliver an exceptionally smooth gaming experience, benefiting gamers even in GPU-limited scenarios.

The Ryzen 7 9800X3D has eight cores and 16 threads that operate at a 4.7 GHz base and 5.2 GHz boost clock rate. The chip employs AMD's 3D V-Cache tech with a new spin, which places a 3D-stacked SRAM chiplet underneath the die to deliver an incredible 96MB of L3 cache to great effect. AMD moved the L3 cache chiplet from the top to the bottom of the compute die this generation. That gives the integrated heat spreader (IHS) direct access to the compute die, allowing for more thermal headroom, and in turn, higher clock speeds. The end result is a comparatively low-power chip that delivers incredible gaming performance and comparable productivity performance to other eight-core models on the market.

3D V-Cache previously came with trade-offs in the productivity department, but that’s not the case with the Ryzen 7 9800X3D. Still, 3D V-Cache doesn’t provide a performance benefit in every game, and the performance benefit is less pronounced as your display resolution climbs.

The Ryzen 7 9800X3D has much lower power consumption than the Intel competition, making it a far cooler processor that won't require as expensive accommodations, like a beefy cooler, motherboard, and power supply. It also takes particularly well to undervolting, which is easy to accomplish with AMD’s Curve Optimizer. That means the 9800X3D delivers top-notch gaming performance and a cooler, quieter, and less expensive system than you'll get with an Ultra 9 or Core i9.

Read: AMD Ryzen 7 9800X3D Review

The $305 Ryzen 7 9700X had a rough initial product launch, but AMD's targeted firmware and operating system improvements have changed the picture tremendously, allowing the chip to place much higher on our CPU benchmark hierarchy (head there for the most up-to-date gaming benchmarks). Combined with lower-than-launch pricing, the Ryzen 7 9700X is a strong contender, tying Intel's Core i9-14900K in gaming and beating the Core i7-14700K. That's not to mention that it beats Intel's entire lineup of Arrow Lake processors as well. Now, all of those processors offer faster performance in heavily-threaded productivity applications than the 9700X, but when it comes to a pure gaming experience, the 9700X either ties or beats all current Intel competitors.

The Ryzen 7 9700X has eight Zen 5 cores with 16 threads that operate at a 3.8 GHz base and 5.5 GHz boost clock. The chip has a 65W TDP, though AMD retroactively added a 105W TDP option you can select in the BIOS that helps boost performance in productivity applications. It's covered by the warranty, as well. With either setting, the 9700X has comparatively tame power consumption, so it is an easy chip to cool. You'll have to buy your own cooler for the processor, though.

The Ryzen 7 9700X drops into socket AM5 motherboards, and B-series motherboards make the most sense for this class of chip. B850 and B840 motherboards get AMD's latest chipset with features like mandatory PCIe 5.0 support on the top M.2 slot and better availability for features like Wi-Fi 7. However, the Ryzen 7 9700X will still work with the older B650 chipset if you can find a board on sale.

Read: Ryzen 5 9700X Review

The $350 Core Ultra 7 270K Plus performs like a flagship CPU, but it costs about half as much. In games, it narrowly outclasses the Core i7-14700K and offers a 2.4% boost over the competing Ryzen 7 9700X. AMD’s last-gen Ryzen 7 7800X3D still offers around a 10% boost over the Core Ultra 7 270K Plus, but it’s also around $50 to $80 more expensive depending on sales.

It’s a solid gaming CPU, and certainly a better recommendation than the Core i7-14700K given prices right now. Compared to the Ryzen 7 9700X, things are tighter. The Core Ultra 7 270K Plus gains an edge with productivity performance. Short of the 9950X, it’s at the top of our multithreaded performance rankings, more than doubling the performance of the Ryzen 7 9700X.

On the gaming front, it supports Intel’s new Binary Optimization Tool, which offers an average of an 8% improvement in gaming performance based on our testing. It’s only available in a limited number of games at the moment, but Intel says it plans to support the feature with updates in the future.

For specs, the Core Ultra 7 270K Plus is close to the 285K. It comes with 24 cores and threads, split across eight Lion Cove P-cores and 16 Darkmont E-cores. The P-cores boost up to 5.4 GHz and the E-cores can climb to 4.7 GHz. Across the CPU, you get a total of 76 MB of combined L2 and L3 cache. It comes with a 125W TDP and 250W MTP. Critically, the Core Ultra 270K Plus also comes with a 900 MHz boost in die-to-die frequency and 400 MHz boost in fabric frequency compared to stock Arrow Lake chips.

The Core Ultra 7 270K Plus slots into existing 800-series motherboards with the LGA 1851 socket. This is an unlocked chip, so if you want to get the full benefits of overclocking, you’ll need a Z890 board. However, it’ll still work with H- and B-series motherboards, just without CPU overclocking support.

Read: Core Ultra 7 270K Plus Review

Mid-Range Best CPU for Gaming - $200 to $300

Intel has returned to gaming prominence with its Arrow Lake Refresh CPUs, and nowhere is that clearer than with the $220 Core Ultra 5 250K Plus. It’s priced like a budget CPU at $220, but it can perform as well (and sometimes even better) than chips that cost twice as much. It doesn’t dominate the gaming charts in the same way as AMD’s X3D offerings, but at this price, it doesn’t need to. It offers marginally better performance than AMD’s competing six-core Ryzen 5 9600X in games while running the tables with application performance.

On average, the Core Ultra 5 250K Plus is 1% faster than the Ryzen 5 9600X at 1080p, and 9% faster than the 245K. It’s functionally identical, but Intel’s new iBOT feature allows the chip to hold some solid leads in certain titles. For instance, it’s 10% ahead of the 9600X in Cyberpunk 2077. Even in a non-iBOT title like Doom: The Dark Ages, the 250K Plus leads by 12%. There are still some games that struggle with the unique Arrow Lake architecture like F1 2024, but the losses are less pronounced with the souped-up Arrow Lake Refresh chips compared to the stock offerings.

The application performance is what really stands out with the 250K Plus, however. With 18 cores, it outpaces the Core i7-13700K, nearly matches the Core i7-14700K, and more than doubles the performance of the Ryzen 5 9600X in multithreaded applications. In single-threaded applications, it beats the Ryzen 5 9600X by 6%.

Although you get 18 cores, they’re split between six Lion Cove performance cores and 12 Darkmont efficient cores. The P-cores climb up to 5.3 GHz, while the E-cores top out at 4.6 GHz. The CPU comes with a combined 60 MB of L2 and L3 cache, along with a TDP of 125W and a MTP of 159W. Like all Arrow Lake chips, it doesn’t support Hyper-Threading, so you get 18 total threads.

The Core Ultra 5 250K Plus slots into existing motherboards with the LGA 1851 socket. It’s unlocked for overclocking, so a Z-series motherboard is an ideal pairing. However, Intel increased the die-to-die frequency and the fabric frequency out of the box, and you’ll see those improvements in action on B- and H-series motherboards, as well. It’s locked to DDR5 memory, unlike Raptor Lake and Alder Lake platforms, and it officially supports speeds up to 7200MT/s.

Read: Core Ultra 5 250K Plus Review

The $230 Ryzen 5 7600X3D is currently the best value gaming CPU you can get right now, though it trades performance in other areas to reach that status. It's just 4.5% slower than the Ryzen 7 7800X3D based on our testing, giving you most of the performance of AMD's coveted 3D V-Cache in games without the extra cost.

In games, it outclasses more expensive CPUs with ease, including the Ryzen 7 9700X, and averaged just 65W of power draw during our gaming tests. Outside of games, however, the Ryzen 5 7600 X3D struggles. The Core Ultra 5 250K Plus is more than twice as fast in multithreaded performance, and in single-threaded performance, even the base Ryzen 5 7600X is around 13% faster.

The lagging productivity performance makes sense. The Ryzen 5 7600X3D is a six-core / 12-thread chip, so it has limited multithreaded potential, and it only clocks up to 4.7 GHz. The limited specs give AMD room to cram 102 MB of combined L2/L3 cache on the die, however, which comes with a sizeable boost in gaming performance. Compared to the base Ryzen 5 7600X, the X3D version is 22% faster despite coming in at lower peak clocks and power draw.

You can slot the Ryzen 5 7600X3D into socket AM5, which is available on 600- and 800-series motherboards, though the latter may require a BIOS update. Memory and CPU overclocking is available on both B- and X-series chipsets; however, the Ryzen 5 7600X3D has a locked multiplier, so the only overclocking you can access is through AMD's Precision Boost Overdrive, or PBO.

Read: Ryzen 5 7600X3D review

Highest Performance Best CPU for Gaming - $400+

How do you improve upon a CPU that already claims a dominating position in gaming and productivity workloads? You add more cache, of course. The Ryzen 9 9950X3D2 is powerful, expensive, and hungry for wattage, but it’s the best of the best if you want top-shelf gaming and application performance. It throws value out the window, and it’s only marginally better than the Ryzen 9 9950X3D, but it is still better.

Based on our testing, it’s about 3.9% ahead of the Ryzen 9 9950X3D in multithreaded applications, and in lockstep in gaming at 1080p. Compared to Intel’s Core Ultra 7 270K Plus, the Ryzen 9 9950X3D2 is 9% ahead in multithreaded performance and 23% ahead in average gaming performance. The Ryzen 7 9800X3D delivers a better value on the gaming front, and the Core Ultra 7 270K Plus is a monster productivity chip at a third of the price of the 9950X3D2. But the magic trick of this chip is that it can do both without breaking a sweat.

Under the hood, the Ryzen 9 9950X3D2 is similar to the Ryzen 9 9950X3D. It’s a 16-core / 32-thread chip packing AMD’s Zen 5 architecture, and it tops out with a 5.6 GHz boost clock; just 100MHz behind the 9950X3D. As the name suggests, this processor is unique because it uses AMD’s 3D V-Cache on both CCDs. Both eight-core CCDs have 32 MB of onboard cache, plus an additional 64 MB chunk placed under the cores, giving you a total of 192 MB of L3 cache.

The extra cache slightly accelerates multithreaded performance overall, though only by around 4%. There are specific workloads where the advantage is more present, with some data science workloads showing performance gains in the realm of 26% over the Ryzen 9 9950X3D. Those specific workstation-class workloads are where the Ryzen 9 9950X3D2 earns its stripes.

Otherwise, it’s the chip to buy because you simply want the best, no matter what the cost or how marginal the improvements are. It slots into existing AM5 motherboards, and it’s best suited for newer 800-series chipsets. AMD officially supports memory speeds up to DDR5-5600, though we find that DDR5-6000 is the sweet spot for Zen 5 CPUs.

Prices have dropped since release, though the 9950X3D2 is still expensive. It launched at $1,000, but you can find the chip for around $900 now.

Read: AMD Ryzen 9 9950X3D2 Review

The prior-gen $469 Core i9-14900K is now selling for all-time low pricing, primarily because the newer $560 Core Ultra 9 285K has arrived to take its place. However, the Core Ultra 9 285K is actually slower than the 14900K in gaming, so it isn't a suitable replacement. The new Core Ultra 7 270K Plus is marginally slower based on our testing, as well. Even in the face of Intel’s 200S Boost update, which was meant improve gaming performance, the competitive landscape remains unchanged. In our testing, the Core Ultra 9 285K gained an average of 7% from the update, which means it’s still slower than the Core i9-14900K.

You should be aware that the much more economically-priced 14700K (listed above) is only 2% slower than the 14900K in gaming but costs over $100 less. The Ryzen 7 9700X, also listed above, is also less expensive and effectively ties the 14900K in gaming.

However, there are Intel fans willing to pay extra for the absolute most gaming performance they can get from an Intel platform. Also, the 14900K does offer more multi-threaded horsepower than the 14700K and 9700X, which could be useful if you game, stream, and record simultaneously or do other heavy multi-tasking while gaming. Just make sure that your use case justifies the extra cost.

The 14900K sports leading-edge connectivity, supporting DDR4-3200 or up to DDR5-5600 memory, along with 16 lanes of PCIe 5.0 and an additional four lanes of PCIe 4.0 from the chip for M.2 SSDs.

The chip comes with eight P-cores that support Hyper-Threading and 16 single-threaded E-cores for a total of 32 threads. The P-cores have a 3.2 GHz base, and peak frequencies reach an amazing 6.0 GHz with Turbo Boost Max 3.0 (this feature is only active on P-cores). Meanwhile, the E-cores have a 2.4 GHz base and stretch up to 4.4 GHz via the standard Turbo Boost 2.0 algorithms. The chip also has 36MB of L3 cache and 32MB of L2.

This 14900K has a 125W PBP (base) and 253W MTP (peak) power rating, but we recorded considerably lower power consumption than its prior-gen counterpart. You'll need to buy a capable cooler for the chip, and you'll also need either a 700-series or 600-series motherboard. Like other Raptor Lake Refresh chips, you can find DDR4 and DDR5 motherboards, though you’ll need to go with a DDR5 board for the highest performance.

The lower price of DDR4 might entice some gamers, but you'll lose anywhere from 5-8% of gaming performance with higher-end Intel chips. You can step up to the much more expensive DDR5 if you need access to more memory throughput and, thus, every bit of performance possible.

Beyond specs, the Core i9-14900K was at the center of a years-long controversy concerning instability. An error in the microcode (CPU firmware) meant the Core i9-14900K would degrade faster than expected, starting with instability in games. Intel has rectified the issue with microcode 0x12F, so make sure you update your BIOS immediately if you pick up Intel’s last-gen flagship

Read: Intel Core i9-14900K Review

Best Budget CPU Pick - $100 to $150

The $164 Ryzen 5 7600X is an attractive budget CPU at its new price, forced down by Intel's new Arrow Lake Refresh chips. It’s marginally slower than the Ryzen 5 9600X, but also marginally cheaper – the Ryzen 5 7600X offers about 90% of the performance of the Ryzen 5 9600X for 94% of the price. It’s a slightly worse value, but it’s still a good option to keep in mind, especially if you find it on sale. The Ryzen 5 7600, sans X, is available at around the same price. We’ve yet to see it drop below the Ryzen 5 7600X, however.

With the 7600X, you get six cores and 12 threads based on the Zen 4 architecture, clocked at 4.7GHz with boost speeds up to 5.3GHz. Unlike the Ryzen 5 9600X, the Zen 4-based version comes with a TDP of 105W. Cooling it shouldn’t be an issue, and you’re free to run in AMD’s 65W Eco mode through the Ryzen Master software.

The Ryzen 5 7600X slots into AM5 motherboards, including 600- and 800-series chipsets, and it supports PCIe 5.0. DDR5 is required, which is a tough pill to swallow at this bang-for-your-buck price point, but it’s hard to avoid soaring RAM prices.

Read: Ryzen 5 7600X review

The $135 AMD Ryzen 5 5600 delivers a solid blend of performance in both gaming and productivity applications, bringing a new level of value to the Zen 3 lineup. If you're fine sticking with a previous-gen AM4 motherboard, the Ryzen 5 5600 makes a great base for a budget build. The primary trade-off for the AM4 platform is that you're limited to DDR4, and you don't have access to PCIe 5.0. You also have a limited runway for upgrades, as the fastest gaming CPUs on AM4 – the Ryzen 7 5700X3D and Ryzen 7 5800X3D – have reached end of life.

The Ryzen 5 5600 also makes an absolutely unbeatable budget chip if you're updating a first-gen Ryzen system. The 5600 unseats the Ryzen 5 5600X, a long-time favorite. The 5600X is only a mostly imperceptible ~1% faster in gaming and multi-threaded PC work than the non-X model, but provides a 4% advantage in single-threaded work.

Our testing shows that the Ryzen 5 5600 generally matches the gaming performance of its more expensive sibling, the ~$230 Ryzen 7 5800X. That makes the 5600 an incredibly well-rounded chip that can handle gaming well, from competitive-class performance with high refresh rate monitors to multi-tasking gaming workloads like streaming, while also serving up more than enough performance for day-to-day productivity apps. As with all AMD CPUs for gaming, you can fully overclock the chip.

The Ryzen 5 5600 has a 3.7 GHz base and 4.6 GHz boost clock. The chip also has a 65W TDP rating, so it runs cool and quiet. Existing AMD owners with a 500-series motherboard will be happy, as the 5600X drops right into existing 500-, 400-, and 300-series motherboards. If you need a new motherboard to support the chip, AMD's AM4 motherboards are plentiful and relatively affordable, with the B-series lineup offering the best overall value for this class of chip.

Prices for the Ryzen 5 5600 have drifted upward as stock depletes, but that’s offset by platform costs. In addition to low prices on AM4 motherboards, the Ryzen 5 5600 is limited to DDR4. High DDR5 prices are a significant roadblock to opting for a newer chip, as prices continue to surge. So it’s hard to recommend a newer budget CPU, even if it’ll net you higher performance. In the event you already have a kit of DDR5, the $189 Ryzen 5 7600 is a compelling option, and it includes AMD’s Wraith Stealth cooler.

Read: AMD Ryzen 5 5600 Review

Entry-Level Best CPU for Gaming - For gaming on integrated GPUs

The $268 Ryzen 7 8700G, AMD's flagship desktop APU, delivers the fastest socketed performance on the market from integrated graphics, bringing passable 1080p gaming to the desktop PC without a discrete graphics card, but its high price point relegates it to a niche audience.

In contrast, the $191 Ryzen 5 8600G delivers 90% of the 8700G's performance but for ~$80 less, making it a solid alternative for gaming systems that don't use a discrete GPU.

Naturally, you'll have to accept lower fidelity settings and be realistic about which titles can play at 1080p resolution. Still, AMD's Hyper-RX suite of features, which includes in-driver Radeon Super Resolution upscaling tech, frame generation with AMD Fluid Motion Frames (AFMF), Anti-Lag+, and Radeon Boost, helps boost performance at a slight cost to image quality. This new feature set, a first for AMD's iGPUs, is a boon for budget gamers.

The Ryzen 5 8600G has six Zen 4 CPU cores and the RDNA 3 GPU engine with eight CUs. The Ryzen 5 8600G drops into the AM5 platform, with value-focused B650 and A620 motherboards being the obvious best combination. These systems offer a new level of connectivity for AMD's APU processors, which were previously on the aging AM4 platform but require DDR5 memory. That adds some cost, so do a value analysis before selecting this processor. If you're looking for the lowest entry price possible with an APU, the Ryzen 5 5600G listed below slots in as the value alternative.

The Ryzen 7 8600G only supports 16 usable lanes of PCIe 4.0 connectivity, while other processors on the AM5 platform support PCIe 5.0. However, we don't feel this will impact this class of system.

More: AMD Ryzen 7 8700G and Ryzen 5 8600G Review

The Ryzen 5 5600G steps into the arena as the value champ for APUs, which are chips with strong enough integrated graphics that they don't require a discrete GPU for light gaming—just be sure you're willing to accept lowered quality settings.

The Ryzen 5 5600G gives you 96% of the gaming performance on integrated graphics than its more expensive sibling, the Ryzen 7 5700G, but for 25% less cash. Our testing shows that its level of performance makes it the best value APU on the market. As long as you're willing to sacrifice fidelity and resolution and keep your expectations in check, the Ryzen 5 5600G's Vega graphics have surprisingly good performance in gaming.

The 5600G's Vega graphics served up comparatively great 1280x720 gaming across numerous titles in our tests, but options become more restricted at 1080p. Of course, you can get away with 1080p gaming, but you'll need to severely limit the fidelity settings with most titles.

With eight cores and 16 threads that operate at a 3.9 GHz base and boost up to 4.4 GHz, the Ryzen 5 5600G also offers solid performance for its price point in standard desktop PC applications. The chip also comes with a bundled Wraith Stealth cooler, sweetening the value prop, and drops into existing 500-series and some 400-series motherboards, though support on the latter will vary by vendor.

The Ryzen 5 5600G is nearly five years old, and as a result, it’s hard to find it in stock at a reasonable price. AMD updated this model with the Ryzen 5 5600GT in early 2024, which features identical silicon and a slight boost to clock speed, and you’ll generally find it for less at around $150.

If your budget is tight and you're looking to build a system for modest gaming, you should check out our Best Cheap CPU feature. Some of those chips can deliver passable gaming performance without a graphics card, and their prices start at just $55 (£40).

Read: AMD Ryzen 5 5600G Review

• MORE: CPU Benchmark Hierarchy
• MORE: AMD vs Intel
• MORE: Best Cheap CPUs
• MORE: Best CPUs for Workstations

AMD has provided instructions on how to run DeepSeek’s R1 AI model on AI-accelerated Ryzen AI and Radeon products, making it easy for users to run the new chain-of-thought model on their PCs locally. Several LLMs using R1 are compatible with RX 7000 series desktop GPUs and select Ryzen CPUs with XDNA NPUs. However, they require the optional driver Adrenalin 25.1.1 to run.

The guide has everything AMD users need to get DeepSeek R1 running on their local (supported) machine. LM Studio has a one-click installer tailor-made for Ryzen AI, which is the method AMD users will use to install R1. AMD shows how the application needs to be tuned for its hardware, including a list of the maximum supported LLM parameters.

DeepSeek R1 allegedly has only recently been distilled into "highly capable" smaller models, small enough to run on consumer-based hardware. The DeepSeek-V3 model was initially trained on a cluster of 2,048 Nvidia H800 GPUs for context.

The maximum supported LLM parameters are based on memory capacity. The RX 7600 XT, 7700 XT, 7800 XT, 7900 GRE, and 7900 XT all support up to “DeepSeek-R1-Distill-Qwen-14B”. The flagship RX 7900 XTX supports up to “DeepSeek-R1-Distill-Qwen-32B”. The RX 7600, with its 8GB of VRAM, supports up to “DeepSeek-R1-Distill-Llama-8B”.

Similarly, Ryzen 8040 and 7040 series mobile APUs are equipped with 32GB of RAM, and the Ryzen AI HX 370 and 365 with 24GB and 32GB of RAM can support up to “DeepSeek-R1-Distill-Llama-14B”. The Ryzen AI Max+ 395 can support up to “DeepSeek-R1-Distill-Llama-70B”, but only in 128GB and 64GB memory capacities. The 32GB supports up to “DeepSeek-R1-Distill-Qwen-32B”.

DeepSeek’s new AI model has taken the world by storm, with its 11 times lower computing cost than leading-edge models. Two days ago, it was solely responsible for Nvidia’s record-breaking $589 billion market cap loss. The DeepSeek R1 model relies on extreme optimization levels to provide its 11X efficiency uplift, relying on Nvidia’s assembly-like Parallel Thread Execution (PTX) programming for most of the performance uplift.

Nvidia and AMD GPUs aren’t the only GPUs that can run R1; Huawei has already implemented DeepSeek support into its Ascend AI GPUs, enabling performant AI execution on homegrown Chinese hardware.

Intel's Arrow Lake-H has been put through its paces by hardware sleuth Golden Pig Upgrade Pack in his latest review at Bilibili. The tested models include the 16-core Core Ultra 9 285H and the 14-core Core Ultra 5 225H, though most of the spotlight is on the flagship variant.

Arrow Lake-H differs slightly from its desktop counterpart because it brings back LPE cores. Since Intel didn't offer much clarification, we aren't sure whether these cores are based on Skymont or Crestmont. In the latter case, Arrow Lake-H likely reuses the SoC Tile from Meteor Lake-H, but let's not jump to conclusions. The P and E cores employ the Lion Cove and Skymont architectures, respectively, cutting off Hyper-Threading.

The Core Ultra 9 285H offers 16 cores (six P-cores + eight E-cores + two LPE-cores) and 16 threads with a turbo clock of 5.4 GHz. The more mainstream-oriented 225H drops to 14 cores (four P-cores + eight E-cores + two LPE-cores), featuring a still-impressive 4.9 GHz turbo frequency. These processors have been compared against the Ryzen AI 9 365 packed with 10 cores (four Zen 5 + six Zen 5c) / 20 threads, and the high-end Ryzen AI 9 HX 370 leading the pack with 12 cores (four Zen 5 + eight Zen 5c) and 24 threads.

The shared images suggest this review was conducted on Lenovo's most recent IdeaPad series outfitted with Arrow Lake. Jumping into the benchmarks, the Core Ultra 9 285H surges ahead of the Cinebench R23 and R24 competition. In the single-core test, the 285H leads its predecessor by 13% with an impressive 26% uplift moving over to multi-core performance. However, all that glitters is not gold, and glancing at the efficiency slide doesn't paint a rosy picture for the 285H.

Limited to just 50W of power, Intel's Arrow Lake-H flagship fails to overtake the Ryzen AI 9 365 and considerably trails the Ryzen AI 9 HX 370. This is quite interesting as desktop Arrow Lake was much faster than Zen 5 in power-limited scenarios. The lackluster efficiency display could be attributed to the testing conditions or the supposedly older SoC Tile, but that's just a guess.

Arrow Lake-H's integrated GPU (iGPU) is based on a modified version of Alchemist with XMX cores. With that in mind, Intel has significantly improved 3DMark's testing suite. The most considerable uplift comes in ray tracing, where Arrow Lake leads Meteor Lake by almost 88%.

The same cannot be said for games where AMD's Radeon 880M is consistently ahead, and by some margin. Here, Alchemist+ yields little to no benefit over standard Alchemist. Funnily enough, Xe2 on Lunar Lake is the polar opposite as it struggles in synthetics but has managed to solidify its position as the fastest gaming iGPU, beating even the Radeon 890M.

After the switch to a disaggregated design, Intel will probably be leveraging the same CPU Tile from the desktop for Arrow Lake-H on mobile. This was an interesting faceoff, but it raises several questions, especially regarding efficiency. Intel didn't specifically discuss Arrow Lake-H's architecture at its keynote, so we can only guess now.

Mini PC maker GMK has been showing off its new EVO-X1 mini PC. It looks like a great little device with one of the latest 'Strix Point' processors from AMD, and plenty of fast ports. However, it also hints that this machine is earmarked for a powerful Ryzen AI Max+ 395 'Strix Halo' processor upgrade in the not-too-distant future.

GMK introduced the EVO-X1 last year — a mini-PC equipped with a Ryzen AI 9 HX 370 chip and Radeon 890M integrated graphics. This is one of the most powerful chips in the current AMD Strix Point lineup, with the CPU boasting 12 cores and 24 threads and the integrated GPU having 16 compute units. The company also gave the EVO-X1 a standard 32GB of memory and a choice between 1TB or 2TB of storage.

What also makes this small computer stand out from the crowd is that it’s the first Ryzen AI 9 HX 370 mini-PC that has an OCuLink port, allowing it to attach to a suitable external GPU without crippling its bandwidth too much. GMK doesn’t have an external GPU enclosure listed on its US website, though, so you’ll have to purchase them separately from other manufacturers, like the Aoostar AG01 or the Minisforum MGA1. But whatever the case, this OCuLink port would allow you to pair the excellent performance of the Ryzen AI 9 HX 370 with a more powerful GPU, up to an RTX 4090 (we wouldn’t recommend that, though, as OCuLink could show an up to 23% drop in performance for high-end GPUs). Connecting with OCuLink should provide much better bandwidth than the Thunderbolt 3 or 4, or USB4 alternatives.

Aside from the OCuLink port (which, unfortunately, isn’t hot-swappable), the EVO-X1 also has two USB-A ports, a USB-C port, and a 3.5 mm combo jack up front, plus the power button right beside the jack. It’s unclear what other USB ports are available at the back, but we also have one HDMI 2.1 and one DisplayPort 2.1 for connecting your display to the onboard graphics. All these features are available for $1,299 for the 1 TB version (although it’s currently on sale at $919.99), while the 2 TB version is priced at $1,599 (with a sale price of $969.99).

Although these specifications are already impressive, GMK is going even further, saying in a press release on Chinese social media site BiliBili (machine translated) that the company “took the lead in launching the Ryzen AI Max+ 395.”

Although details are scant at the moment, the company shared a Geekbench result for the Vulkan test, which showed the processor, presumably on an EVO-X1 mini-PC evaluation platform, achieving 67,004 points. This score beats the RTX 3050 and RX 5700 desktop graphics cards, showing how far integrated GPUs have come along and the potential of AMD’s top-end Strix Halo chip for its mini-PC.

Still, you should note that reviewers rarely use Geekbench 6 to measure GPU performance and that we should wait for independent reviews to see what this mini-PC could do. Whatever the case, we want to see how the EVO-X1 mini-PC with a Ryzen AI Max+ 395 would stack up as a gaming and productivity machine, and if its built-in GPU could really lead to the end of entry-level graphics cards.

Razer is honing its gaming laptop lineup with an all-new Blade 16 for 2025. Not only does Razer claim that the next-generation Blade 16 is the thinnest laptop that it's ever produced, but it also comes packed with potent hardware for gaming enthusiasts.

Starting with the design, Razer has crafted a new chassis for the Blade 16, which is just 0.59 inches at its thinnest point. As before, the Blade 16 employs an aluminum unibody to maximize strength, and the vapor chamber cooling system has been updated to improve the performance of the dual cooling fans. This should lead to lower chassis temperatures and ensure that fan noise is minimized during intense gaming sessions.

Speaking of gaming, the Blade 16 has the chops to put it among the laptop gaming elite, with an AMD Ryzen 9 HX 370 onboard paid with LPDDR5X-8000 memory. It is a 12-core/24-thread processor with a maximum boost clock of 5.1 GHz. In addition, the processor has an integrated 50 TOPS neural processing unit (NPU), meaning that it’s fully compliant with Microsoft’s Copilot+ functionality in Windows 11.

The move to AMD is notable, as Intel processors have traditionally powered the Blade 16. Intel has been taking it on the chin for quite some time in the hardware space, and losing out in a high-profile system like the Blade 16 shows how much the company has to prove with its new mobile-centric Arrow Lake processors.

On the graphics front, Razer spared no expense. You’ll find Nvidia’s newly announced GeForce RTX 50 Series graphics cards, including the flagship GeForce RTX 5090 laptop GPU with 24GB of GDDR7 memory. Nvidia is promising some substantial performance gains over the preceding GeForce RTX 40 Series, aided by the artificial intelligence (AI) prowess of DLSS 4. Although we know that RTX 5090 (155W) is offered, we don’t know which of the other members of the family — RTX 4080, RTX 4070 Ti, RTX 4070 — will also be available on the Razer Blade 16.

All that CPU and GPU horsepower would be for naught if it connected to a crummy display. Thankfully, Razer is sticking to high-end hardware by fitting the Blade 16 with a 16-inch QHD+ OLED display with a 240 Hz refresh rate. It has a 0.2 ms response time, and the bezels measure in at a thin 4.7 mm.

Other niceties include an all-new keyboard with 1.5 mm of key travel, five customizable macro keys, and a dedicated Copilot+ key. There’s also a new six-speaker audio system with THX tuning.

Razer says the all-new Blade 16 will be available in Q1 2025, but pricing has not yet been announced. However, given that the previous generation started at $2,999, expect to pay at least that much for this RTX 50-equipped evolution.

OneXPlayer has officially launched the X1 Pro, featuring a 3-in-1 form factor powered by AMD's Strix Point APUs. Following the F1 Pro and the Ayaneo 3, the X1 Pro is one of the first handhelds that employs AMD's Zen 5 Strix Point APUs and is a strong competitor to GPD's forthcoming Pocket 4 mini-laptops.

The OneXPlayer X1 Pro has a multi-purpose design, as it can simultaneously be used as a laptop, a tablet, and a handheld. Just like the Pocket 4, the X1 Pro offers the Ryzen AI 9 HX 370, which features 12 cores (four Zen 5 + eight Zen5c) in a dual-CCX layout, 12MB of L2 cache alongside 24MB of L3 cache, and a boost clock of 5.1 GHz. In addition, on the iGPU (Integrated GPU) side comes the Radeon 890M with 16 CUs (Compute Units). Powered by AMD's XDNA 2, the NPU on the HX 370 can dish out 50 TOPS of AI performance.

The display is decked out with a 10.95-inch 120 Hz LTPS (likely IPS) panel with an impressive 2560x1600 (2.5K) resolution. The X1 Pro is also vivid, as the panel can cover 138% of the sRGB color space with a brightness of 540 nits. You may configure the storage to your liking, from 1TB to 4TB, with speeds of up to 7.1 GB/s (read) and 6.5 GB/s (write) in the 2280 form factor. Moreover, OneXPlayer has gone one step ahead and included a MicroSD card slot for additional storage.

The X1 Pro's 65.02Wh battery promises eight hours of battery life, though real-world runtimes will vary; when configured at 28W, it should last around two hours. OneXPlayer has also included a 100W Gallium Nitride charger in the box. Connectivity-wise, there is support for Wi-Fi 6E and Bluetooth 5.2.

The added controller features hall-effect RGB joysticks to ward off stick drift. Regarding I/O, we see two USB 4.0 ports, an OCuLink interface, a MicroSD card slot, a USB 3.2 Type-A port, and a 3.5mm audio jack.

Pre-orders are live starting today, and you can snag the X1 Pro at a discounted price while the presale lasts. Shipping is expected to start on January 3rd, and OneXPlayer will demo the X1 Pro alongside other devices at CES 2025 next month.

The new high-end PC handhelds are imminent— the presale window of OneXPlayer's OneXFly F1 Pro handheld, the world's first AMD Ryzen AI 300-based handheld gaming PC, is here and open through the end of November per the official OneXPlayer store listing.

The OneXFly F1 Pro should give the Asus ROG Ally X and its ilk of Ryzen Z1 Extreme/Ryzen 7 8840U, particularly its Ryzen AI 9 HX 370 model. However, even the Ryzen AI 9 365 model can outperform the last generation while sparring Watt-to-Watt, which we'll explain in more detail soon.

For now, prospective buyers should know that of the six OneXFly F1 Pro configurations available from OneXPlayer in presale, there are only two models—one with Ryzen AI 9 365 and one with Ryzen AI 9 HX 370. The two Ryzen AI 9 365 models differ only in storage capacity (1 TB or 2 TB) and start at $1,099, while the four Ryzen AI 9 HX 370 models scale from 1 TB to 4 TB storage and 32 GB to 64 GB RAM and start at $1,339.

*Note: Ryzen AI 9 365's Zen 5 configuration has 4 Zen 5 cores and 6 Zen 5C cores. Ryzen AI 9 HX 370's Zen 5 configuration is 4 Zen 5 cores and 8 Zen 5C cores.

Beyond raw specifications pointing toward the capacity to provide a significant amp for the coming year of new AMD handhelds, the OneXFly F1 Pro also offers a 7-inch OLED display capable of 800 nits brightness, which should be suitable for HDR and high-contrast scenes in general.

The Steam Deck OLED's screen reaches 1000 nits but is only 90 Hz and operates at a much lower 1280 x 800 resolution. Still, the trade-offs OneXPlayer made here for resolution and refresh rate would be helpful in some titles. At this screen size, resolution scaling can more than compensate for the high-resolution display in more intensive games, particularly with improved FSR support across Windows and Linux alike.

While it is worth noting that the most thorough benchmarking of the OneXFly F1 Pro is currently only available from OneXPlayer themselves, the results provided are perfectly in line with what we expect from the specifications of the handheld. The most impressive gains really might be in the Ryzen AI 9 365 model— not only is it able to outperform its predecessors watt-for-watt, but its Radeon 880M iGPU has the same number of compute units as the last-gen Radeon 780M, highlighting the improvements from RDNA 3 to RDNA 3.5.

Display and performance characteristics considered, the OneXFly F1 Pro quickly becomes a contender for the top dog of the handheld PC market until competitive options with similar internals start popping up. This handheld is only lacking compared to high-end models, such as OCuLink or similar eGPU support. While USB4 ports are nice, they won't quite reach the throughput and latency required for a great external GPU gaming experience. So, using OneXFly F1 Pro as a dockable handheld gaming PC for more power will work, but maybe not to the extent you'd hope at this price point.

Finally, that presale pricing starts at $1,099 or $1,399 depending on CPU— are these performance gains worth all the extra money you'll spend over the Asus ROG Ally X, which starts at $799? Is the Asus ROG Ally X even worth all the extra money you must pay over Steam Deck OLED, which starts at $549 when you have to plug it in to get extra frames? We can't make those value conclusions for you, but we can give you the information you need to decide for yourself. If you want the best handheld PC performance currently available and don't mind the diminishing returns in frames per dollar, we recommend securing a presale OneXFly F1 Pro while you can.

Everyone else should likely wait for further reviews or more competition, though it's worth noting that each model's price will increase once the presale period has ended.

AMD has published a blog post boasting of incredible gaming performance for its Ryzen AI 300 series processors, with performance that is on average 75% faster than Intel's equivalent Lunar Lake counterpart. That number doesn't agree with our own testing of the Radeon 890M and Arc 140V integrated graphics, in part because we didn't use "performance enhancing" settings that are inherently unequal.

AMD unveiled several gaming benchmarks at 1080p medium settings, showcasing the Ryzen AI 9 HX 370 with Radeon 890M significantly outperforming Intel's competing Core Ultra 7 258V with Arc 140V graphics. Games selected for the comparison included Ghost of Tsushima, Cyberpunk 2077, Dying Light 2 Stay Human, Baldur's Gate 3, Spider-Man Remastered, Shadow of the Tomb Raider, Hogwarts Legacy, Assassin's Creed Mirage, Call of Duty: Black Ops 6, Borderlands 3, Doom Eternal, Tiny Tina's Wonderlands, Forza Horizon 5, Hitman 3, Far Cry 6, and F1 24.

On average, AMD touts a 75% performance lead for its Radeon 890M powered Zen 5 mobile chip. For example, in Black Ops 6, the HX 370 generated an average frame rate of 99. The 258V, by contrast, held an equivalent frame rate of just 48. It sounds good, until you look at the specifics of the testing.

AMD is taking advantage of all its software performance boosters to yield these impressive gains. This includes using both FSR 3 upscaling as well as FSR 3 frame generation, and in games that don't support FSR 3, AFMF 2 driver-based frame generation is applied as a substitute. HYPR-RX was also enabled for games that support it, which will enable one or more of: Anti-Lag, AMD Fluid Motion Frames 2 (AFMF2), Radeon Super Resolution (RSR), Radeon Boost, and/or FidelityFX Super Resolution (FSR). RSR and AFMF2 work via AMD's drivers to enable frame generation and spatial upscaling (FSR1) for games that don't have native support upscaling/framegen, while Radeon Boost dynamically lowers the render resolution when fast motion is detected.

For the Intel counterparts, AMD uses Intel's competing XeSS solution for comparison, as well as FSR in cases where XeSS isn't present. But it's important to note that XeSS doesn't have any type of frame generation functionality, which inevitably skews the results in favor of AMD. Also, we've seen in the past the FSR3 runs better on AMD's GPUs than on competitors' GPUs, which makes sense, but also results in non-equal workloads.

AMD also showcased the native performance of both processors without any upscale or frame generation enabled. In this software environment, the AMD Radeon 890M-equipped CPU was pushing virtually identical frame rates, on average, compared to the Arc 140V-equipped Intel counterpart. The Core Ultra 7 258V outperformed the Ryzen AI 9 HX 370 in several games, including Black Ops 6 and Hitman 3. But AMD didn't provide any exact framerate statistics for the native performance figures, merely showing a bar graph visual and making it impossible to define clearly how much faster or slower the 258V was against the HX 370.

Like any manufacturer provided benchmarks, take the 75% performance uplift headline claim with a large helping of salt. While gamers can choose to enable such technologies, using them to compare performance can heavily skew the results. For the CPU world, it would be like running Cinebench but only rendering a quarter of the pixels and then using a denoising algorithm versus a competitor that has to fully render every pixel.

What AMD is advertising is a potential 75% performance uplift by using software techniques to reduce the graphics workload and boost performance. But without that extra software, AMD's own numbers show its Ryzen AI 9 HX 370 performs on par with Intel's Core Ultra 7 258V. That mostly matches our own testing, which gave Intel's GPU a 6.7% lead overall at 720p and a 5.5% lead at 1080p.

Famous handheld console maker Ayaneo announced its latest flagship device, the Ayaneo 3. This builds upon the success of its predecessor, retaining a similar shape and button layout but comes with several hardware improvements.

According to the Ayaneo blog, the Ayaneo 3 can be purchased with either an AMD Ryzen 7 8840U processor or the more powerful AMD Ryzen AI 9 HX 370. It also features a seven-inch high-refresh LCD or OLED HDR display, delivering the ultimate immersion. Ayaneo added two buttons at the back of the console to give users more control options.

Unfortunately, Ayaneo hasn’t released additional information yet about the Ayaneo 3. So, we’ll have to wait for further details to learn more about its RAM, storage, and pricing information. Its predecessor, the Ayaneo 2, is currently available with either 16GB or 32GB of RAM and 512GB, 1TB, or 2TB of storage. It also had a starting price of $1,099 and could go as high as $1,549 if you picked the limited-edition B.Duck model. Given that information, it’s likely that we will get the same options for the Ayaneo 3 at a slightly higher price point.

This new addition to the Ayaneo lineup makes it one of the first handheld consoles to utilize the AMD Ryzen AI 9 HX 370. Although most current games do not use the NPU included in the HX 370 processor, its four extra cores, Zen 5 architecture, and AMD Radeon 890M integrated GPU will give it more oomph than the Ryzen 7 8840U processor. Just this processor makes it more powerful than some of the best PC gaming handhelds, which are mostly powered by the Ryzen Z1 Extreme.

The Ayaneo 3 will enter a market already saturated with great options. At the moment, we expect the ROG Ally 2, MSI Claw 8 AI+, and Lenovo Legion Go Gen Two to arrive before the end of the year or early 2025. We want to see how the Ayaneo 3 will differentiate itself from the competition and what kind of experience it will deliver to handheld console gamers.

Handheld maker OneXPlayer uploaded a teaser video of its next gaming device, the Onexfly F1 Pro, on YouTube. It features AMD's Ryzen AI 9 HX 370, part of the Ryzen AI 300 (codenamed Strix Point) family. The new device, dubbed the Onexfly F1 Pro, is a 7-inch handheld that was shown running Black Myth: Wukong at between 50 and 60 FPS.

The new device purportedly features a bleeding-edge HDR-supported 7-inch OLED display with a refresh rate of 144 Hz. It weighs 598 grams and has Harman Kardon speakers. The main selling point is the Ryzen AI 9 HX 370 CPU, which features four Zen 5 cores, eight Zen 5c cores, and AMD's latest Radeon 890M iGPU with 16 CUs based on the RDAN 3.5 architecture. With such robust specifications, the OneXFly F1 Pro can compete against rivals like the Steam Deck OLED, ROG Ally X, or Lenovo Legion Go.

OneXPlayer showcased the device running the Black Myth: Wukong benchmark at 1080p at low-quality settings with 65% upscaling (1248 x 702 effective internal rendering resolution). The AI 9 HX 370-equipped handheld generated an average of 58 FPS at these settings. Power consumption was limited to 15W for this test.

The Onexfly F1 Pro will be OneXPlayer's first device to sport both an OLED display and AMD's latest Zen 5-based mobile CPUs; all its current devices sport previous-gen Intel or AMD CPUs along with non-OLED-based displays. It will also be one of the first Zen 5-powered handhelds to hit the market, with the GPT Pocket 4 being the only other apparent device advertised with the Ryzen AI HX 370.

Compared to the Pocket 4, the F1 Pro is a true handheld with traditional grips and controls, while the Pocket 4 is a hybrid 2-in-1 device with a keyboard and a screen that flips 180 degrees.

OneXPlayer's benchmark teaser demonstrates that the AI 9 HX 370 is ideally suited to handheld devices and can play the most demanding AAA titles even at a mid-range 15W TDP to extend battery life. Since AMD has yet to launch its next-generation handheld-focused Z-series CPUs, handheld makers are now stuck using AMD's laptop-focused Ryzen AI 300 series CPUs.

AMD claims that its Ryzen AI 300 (codenamed Strix Point) offerings can easily beat Intel's latest Core Ultra 200V (codenamed Lunar Lake) CPUs in consumer LLM workloads. Team Red has showcased several charts - comparing the performance of processors from both lineups - with AMD in front by upwards of 27%.

Since the AI boom, tech giants—both on the hardware and software fronts—have entered an arms race to outpace each other in the AI landscape. While much of this is purely for the quote-unquote "AI hype," benefits for mainstream customers are starting to materialize. Microsoft demands a minimum of 40 TOPS for any system to be considered a "Copilot+ PC." Almost every CPU manufacturer now reserves valuable silicon space for a Neural Processing Unit (NPU).

On that note, AMD says that its Strix Point APUs seemingly outperform Intel's Lunar Lake when running LLMs locally. LM Studio's typical consumer LLM is run based on the llama.cpp framework. The Ryzen AI 9 HX 375 leads the Core Ultra 7 258V by up to 27% in the CPU department. The latency benchmarks are an absolute bloodbath for Intel - with AMD reportedly delivering 3.5X lower latency in the Mistral Nemo 2407 12b Instruct model. When switching to the integrated graphics solution, the Ryzen AI 9 HX 375's Radeon 890M iGPU is, at best, 23% faster than Intel's Arc 140V.

Interestingly, while the HX 375 is the fastest Strix Point APU, the Core Ultra 7 258V is not. The flagship Core Ultra 9 288V offers a full 30W TDP, which aligns it relatively better with Strix Point since the latter can be configured as high as 54W.

That aside, AMD then tested both laptops in Intel's own and first-party AI Playground - where we see AMD in front by 13.1% using the Mistral 7b Instruct v0.3 model. The Microsoft Phi 3.1 Mini Instruct test is slightly less exciting as the lead drops to 8.7%.

All in all, it is not every day that you fire up your laptop to run a local LLM, but who knows what the future holds? It would've been more exciting to see the kingpin Core Ultra 9 288V in action, but don't expect drastically different results. AMD says that it is actively working to make LLMs, currently gated behind many technical barriers, more accessible to everyone, which highlights the importance of projects like LM Studio.

It’s an exciting time to be in the market for a new ultraportable laptop. Qualcomm wants to make waves with its Snapdragon X Series of Arm processors, which boast incredible battery life. Intel just launched its new Core Ultra 200 “Lunar Lake” chips, which have a large focus on overall efficiency. And, of course, AMD is pushing its frontline Ryzen AI 300 processors in the laptop space.

The HP OmniBook Ultra review unit we tested comes from the latter series of chips; specifically, it features a Ryzen AI 9 HX 375 processor. That powerful mobile chip is ensconced within an attractive and lightweight aluminum chassis. The design is understated, with the only bold move being a light blue power button that doubles as a fingerprint reader.

The OmniBook Ultra proved to be a swift performer in our benchmark suite, putting it in contention among the best ultrabooks.

Design of the HP OmniBook Ultra

Less is more when it comes to the OmniBook. The laptop doesn’t draw attention to itself, which suits those who like to fly under the radar sans blaring.

The lid is simple and finished in aluminum with the HP logo polished to a mirror finish. Opening the lid reveals a no-nonsense keyboard deck flanked by two rectangular-shaped speaker grilles. The keyboard is dark gray in finish, while the top function row is a lighter shade of gray. The only design embellishment is the power button, which is finished in a pale blue and doubles as a fingerprint reader.

The display is covered in glossy glass, which is a boon for increasing color fidelity, but amplifies reflections. The bezels are minimal along the sides, but are more sizable for the top and bottom (where you’ll find another HP logo).

There are also two Copilot logos on the chassis. One is located at the base of the display lid, and is only visible when the lid is closed. The second is located under the left CTRL key on the keyboard. And if you didn’t already know that this machine was powered by AMD, two stickers are placed on the right palm rest to advertise the Ryzen AI 9 HX processor and Radeon Graphics inside.

Regarding external ports, the OmniBook Ultra doesn’t include many. There are two Thunderbolt 4 ports on the right side of the chassis (a huge rarity for AMD systems, as Intel licenses Thunderbolt 4), while a 3.5 mm combo jack and a USB-A port are on the left side. However, the USB-A port has a fussy, spring-loaded half-cover that mirrors the contour of the laptop. This design is often included with laptops with a built-in Ethernet port, but I’ve never seen it for a USB-A port before. It seems like an unnecessary part that could become easily broken.

The OmniBook Ultra measures 12.41 x 8.96 x 0.65 inches and weighs 3.47 pounds. That is slightly heavier than the larger Asus ZenBook S16, which weighs 3.31 pounds and measures 13.92 x 9.57 x 0.47 inches. The ZenBook S14 has dimensions of 12.22 x 8.45 x 0.51 inches and weighs 2.65 pounds, while the Dell XPS 13 (9345) measures 11.63 x 7.84 x 0.60 inches and weighs 2.62 pounds.

HP OmniBook Ultra Specifications

Productivity Performance on the HP OmniBook Ultra

This is the first laptop we’ve tested using AMD’s new Ryzen AI 9 HX 375, which sits just above the Ryzen AI 9 HX 370 we previously reviewed in the Asus ZenBook S16. The processor features four Zen 5 cores and eight Zen 5c cores. The "Strix Point" chip supports a total of 24 threads. The Zen 5 and Zen 5c cores each have a base clock of 2 GHz but boost to a maximum of 5.1 GHz and 3.3 GHz, respectively.

These specs are all identical to the Ryzen AI 9 HX 370. The only difference between the two chips is that the Ryzen AI 9 HX 375 has a slightly faster XDNA2 NPU, which boosts maximum AI performance from 50 TOPS to 55 TOPS.

Our OmniBook review unit came configured with 32GB of LPDDR5x-7500 RAM and a 2TB M.2 PCIe NVMe SSD.

Kicking things off with the synthetic Geekbench 6 benchmark, the OmniBook Ultra flexed its muscles with wins in single-core (2,847) and multi-core (14,838) performance. The XPS 13 (Snapdragon X Elite) was nipping at its heels, scoring 2,797 and 14,635, respectively. Rounding out the slate of competitors was the ZenBook S16 (Ryzen AI 9 HX 370) at 2,765/13,282 and the ZenBook S14 (Core 7 Ultra 258V) at 2,751/11,157.

The OmniBook Ultra had another convincing win in our file transfer test, where it copied 25GB of files at 1,708.23 Mbps. The next-closest competitor was the XPS 13 at 1,342 Mbps, while the ZenBook S16 lagged well behind in last place with 908.45 Mbps.

The dominance of HP’s new OmniBook Ultra extended to our Handbrake test, which transcodes a 4K video file to 1080p. The OmniBook Ultra completed the task in 4:18, putting it nearly 30 seconds ahead of the second-place XPS 13. The ZenBook S14, with its newly launched Lunar Lake-based processor, took almost twice as long as the OmniBook Ultra to complete the Handbrake test (8:28).

To stress test ultrabooks, we rely on Cinebench 2024, which we run through ten loops. The first run was the highest, at 960.85, while subsequent runs meandered through the mid-900s to low-900s. During the stress test, the Zen 5 cores averaged 3.1 GHz, while the Zen 5c cores averaged 2.37 GHz. During the test, the CPU measured an average of 70 degrees Celsius (158 degrees Fahrenheit).

NPU and AI

The Ryzen AI 9 HX 375 is nearly identical to the Ryzen AI 9 HX 370 in key specifications, apart from the NPU. The NPU in the former receives a 10 percent performance uplift, maxing out at 55 TOPS instead of 50 TOPS. To qualify for Microsoft’s Copilot+ PC certification, the PC must be capable of 40 TOPS, which means that the Ryzen AI 9 HX 370 easily clears the bar.

Copilot+ PCs will use the NPU for features like Cocreate, an AI tool to augment your drawings, and Windows Studio Effects, which provides real-time effects for your camera. However, the biggest headliner for the NPU will be its use with the controversial Recall feature in Windows 11. Recall takes snapshots of everything you’ve done on your computer, which is then accessible by searching for keywords or browsing through a timeline of your PC’s recorded history. Microsoft now intends for Recall to become available for Copilot+ PCs in October via the Windows Insider program.

Display on the HP OmniBook Ultra

While many ultrabooks opt for OLED panels for deeper blacks and richer colors, HP sticks with tried-and-true IPS technology for the OmniBook Ultra. The 14-inch panel has a 16:10 aspect ratio and a relatively uncommon 2240 x 1400 resolution. Our review unit was equipped with the touch panel option, and the refresh rate clocks in at 60 Hz.

The OmniBook Ultra covers 75.6 percent of DCI-P3 and 107 percent of sRGB, putting it about average in this assembled group of laptops. Maximum brightness measured 359.4 nits, slightly edging out the ZenBook S16 (357 nits) for second place. The XPS 13 took top honors at 456 nits.

Touch screens on laptops aren’t a big draw for me, but HP includes the functionality here for those who like that input method. The touch support worked as expected, but the thought of smearing the glossy screen with my fingerprints just gives me the heebie-jeebies.

Keyboard and Touchpad on the HP OmniBook Ultra

I love the look of the OmniBook Ultra’s keyboard. The laptop's chassis resembles a medium graphite color, while the primary keys are dark gray with white lettering. The function row is a lighter shade of gray.

Key presses are firm but very responsive. Although the keys have plenty of travel for my taste, the key presses are a bit on the loud sound. They remind me of older MacBook keyboards when Apple was in its ill-conceived scissor-switch phase. However, the half-height up and down arrow keys are the biggest faux pas. I prefer full-size keys all-around, but some manufacturers are intent on squeezing in half-height keys to save space.

As always, I use keyhero.com to test my typing speed. The test clocked me at 84.78 words per minute with 96.34%accuracy.

The OmniBook Ultra uses a glass touchpad that measures 4.9 x 4.4 inches and has a satisfying click with each press. The touchpad is top-hinged, meaning the presses from the bottom of the touchpad to about three quarters of the way up register with ease. However, pressing in that last quarter of the touchpad’s surface requires a much harder press to register. The surface was smooth, allowing me to easily glide my finger without dragging.

Audio on the HP OmniBook Ultra

HP advertises the OmniBook Ultra as having “pitch perfect audio by Poly,” but in practice, it’s good, but by no means perfect. It has quad speakers with HP Audio Boost and DTS:X Ultra support.

I first queued up a Throttle House video of the Tuthill 911K vs Alfaholics GTA-R 290. What should have been an aural explosion of mechanical amazingness from the 911K’s engine revving to 11,000 rpm sounded merely pedestrian compared to the 14-inch MacBook Pro that was sitting beside the OmniBook Ultra playing the same video.

However, music sounded a bit better, as I could make out some slight inklings of bass with the drum hits in the Fugee’s “Killing Me Softly.” Lauren Hill’s vocals were sharp and clear, with no distortion when pushing the volume towards 100% (which you must do, because this is such a classic 90s banger).

Upgradeability on the HP OmniBook Ultra

The OmniBook Ultra’s bottom cover is affixed with just four Torx screws, which was a relief – I'm used to laptops having eight or more. After removing the screws, a pried the bottom cover off by starting with the exposed lip at the back near the exhaust vents.

At first glance, there doesn’t appear to be anything user-accessible, save for the battery. However, the M.2 slot for the Wi-Fi 7 card is hidden under a small, thin plastic I/O shield held in place with an adhesive. A larger plastic I/O shield is located just below the left cooling fan and heatpipe. Once removed, you have access to the M.2 SSD.

The LPDDR5x memory is soldered onboard, as is the case with many ultrabooks, so there’s no opportunity to perform upgrades post-purchase.

Battery Life on the HP OmniBook Ultra

The OmniBook Ultra was the leader regarding performance. Despite this, battery life didn’t suffer significantly compared to the x86-based competition at 12:52. The ZenBook S14 lasted an hour longer (13:51), but that laptop lagged in all performance metrics. Likewise, the ZenBook S16’s runtime was over an hour shorter (11:35).

The Arm-powered XPS 13 was the outlier, giving up after an impressive 19-and-a-half hours.

Heat on the HP OmniBook Ultra

We measure heat output during the Cinebench 2024 stress test. Despite its thin frame and high performance (among this competition), the OmniBook Ultra was never uncomfortable to touch.

We measured 104.2 degrees Fahrenheit between the G and H keys, while the touchpad was 79.5 F. The hottest part of the laptop was on the bottom between the two cooling fans. The temperature increased to 116 F in this area.

Webcam on the HP OmniBook Ultra

HP uses a 1080p webcam with IR capabilities on the OmniBook, which means that Windows Hello facial recognition is supported (in addition to the built-in fingerprint reader). The images captured by the webcam were “OK” and didn’t blow me away. Fine detail, like my facial hair and patterns on my shirt, looked slightly blurry. In addition, the images/video looked a bit cool. What should have been deep red on my shirt instead looked a slightly paler facsimile.

Software and Warranty on the HP OmniBook Ultra

HP went overboard with the installed software on the OmniBook Ultra. You’ll find HP AI Companion, HP Enhanced Lighting (lighting controls for the webcam), HP Energy Star (an app that just shows you Energy Star appliances and allows you to search for local rebates), HP PC Hardware Diagnostics, HP Privacy Settings, HP Smart (for setting up HP printers), HP Support Assistant, and HP System Event Utility. Would it have hurt HP to put all of these into a single app?

In addition, there’s the Omen Gaming Hub, even though the OmniBook Ultra isn’t explicitly billed as a gaming system. The other preinstalled software includes DTS Sound Unbound, Poly Camera Pro (which gives you more granular control over the webcam), and a link to Otter.ai and McAfee Antivirus. That’s a lot of bloatware for one system.

The OmniBook Ultra ships with a one-year warranty from HP.

HP OmniBook Ultra Configurations

The AMD flavor of the OmniBook Ultra is available in ready-to-ship or customizable configurations directly from HP. The preconfigured system comes with a Ryzen AI 9 365 processor, 16GB of LPDDR5x, and a 1TB PCIe NVMe SSD for $1,079.99.

Our review unit, however, was customized (starting price of $1,399.99) with a Ryzen AI 9 HX 375 processor, 32GB of LPDDR5X, and a 2TB SSD for $1,748.99.

Bottom Line

The HP OmniBook Ultra is an all-around good thin and light laptop. The Ryzen AI 9 HX 375 processor proved a potent motivator, allowing the laptop to take top honors in all the performance benchmarks. Even its storage subsystem performance outshined its rivals.

While you might expect the high-performance chip to impact battery life significantly, the OmniBook Ultra was right in line with its peers in endurance. Only the Dell XPS 13 with the Snapdragon X Elite Arm processor showed a dramatic advantage in battery life.

The display proved to be vibrant, and the speakers were good, but not great. It’s hard to find fault with the hardware other than a keyboard that is a little loud and the unnecessarily fussy USB-A port. The biggest detractor is the sheer amount of bloatware that HP includes with the OmniBook Ultra, but that can luckily be rectified with some time spent in the Installed Apps section of Settings.

At $1,748.99 as-tested, the OmniBook Ultra isn’t cheap, but it’s a high-flying performance powerhouse without a lot of compromises.

MORE: How to Buy a Gaming Laptop

MORE: Best Gaming PCs

MORE: Best Ultrabooks and Premium Laptops

The next big wave of Ryzen AI 300-based mini PCs from Beelink, Aoostar, and Minisforum are set to hit the market in the next few weeks. Yesterday, a leaked graphic revealed the design of the Minisforum EliteMini AI370 ahead of its official reveal. We also received an email from Minisforum giving the full specifications of the upcoming EliteMini AI370.

The core specs of the EliteMini AI370's CPU, RAM, and storage are all in line with what we expect from this category. There's also 16GB of LPDDR5X RAM and 1TB of NVMe PCIe 4 SSD storage. The speed of the SSD is not specified, but the RAM is rated for 7500 MT/s. There's an extra NVMe slot, so total SSD capacity will total 4TB with both slots occupied.

The EliteMini AI370 has a Ryzen AI 9 HX 370, which is a 12-core, 24-thread CPU based on AMD's Zen 5 CPU core design, paired with a Radeon 890M iGPU, which has 16 RDNA 3.5-based Compute Units. By comparison, the best last-gen iGPU used in mini PCs and handhelds — the Radeon 780M — only had 12 of these Compute Units. So, not accounting for architectural or CPU improvements, that's roughly a 30% boost in graphics performance.

Besides that, all you really need to know about is the size and the I/O, both of which are partially pictured above. The real size of this Mini PC measures 5.1 x 5 x 1.85 inches (130 x 127 x 46.9 mm) — sure to fit nicely onto any confined desk setup you have. The I/O is nice, too, with one audio port, one USB 4 port, and two USB 3.0 ports up front. The rear I/O is a little more decked-out with two additional USB 3.0 ports, one HDMI port and one DisplayPort, two Ethernet ports, and an additional audio port.

At this point, the only thing that seems to be missing from this lineup of new Mini PCs seems to be OCuLink support. OCuLink would allow for more external GPU performance than a USB 4 or Thunderbolt 4 connection could ever hope for, but it's still a surprisingly rare feature (despite how many PCs are leveraging these iGPU-focused designs).

Intel officially released its Lunar Lake mobile processors today, and we have a review of the Asus Zenbook S14 with the Core Ultra 7 258V. This is supposed to be the new king of laptop processors, with improved battery life being a key element. But it's also the showcase for Intel's latest GPU architecture, and while no integrated graphics solution will rank among the best graphics cards and these aren't positioned as gaming laptops, we want to see how Intel's latest stacks up against the competition.

The Core Ultra 9 288V comes with Intel Arc Graphics 140V, with the V indicating the power level of the chip. It's a nominal 30W part with a maximum turbo power of 37W. We're using the Core Ultra 9 in order to potentially remove power restrictions that might impact the Core Ultra 7 258V, with the hope of seeing higher, more consistent performance. This is also the debut of Intel's Battlemage GPU architecture, sort of — power limits and shared memory will be major limiting factors for any GPU workload.

To see what Lunar Lake and Battlemage bring to the table, we'll compare the new processor with the previous generation Meteor Lake chip along with AMD's Ryzen AI chip. Here's the quick rundown of core specs for the various chips we're looking at.

[Note: We have retested the games that used FSR3 upscaling and framegen to eliminate those technologies from the equation, as it seems framegen may have failed to work in at least one instance. We also tested with XeSS alongside FSR3 upscaling on both AMD and Intel GPUs, without framegen.]

What should be clear from the above table is that the three laptops we have for testing are not equivalent in a variety of ways. There are slightly different power targets, different CPU core counts, different tiers (Core Ultra 7/9 and Ryzen 9), different memory, and other places where they don't match up. There are also wildly different NPUs in play for AI workloads and even wildly different AI capabilities for the GPUs.

But I'm mostly interested in the graphics performance, so that will be the focus. And even here, we have some interesting changes relative to the previous generation Meteor Lake GPU. Both have eight Xe-cores (roughly similar to an AMD Compute Unit, aka CU), but where Meteor Lake's Arc Xe GPU omitted the XMX functionality, Lunar Lake's Arc Xe2 includes that. It should allow for better XeSS upscaling quality and performance; otherwise, we don't expect that to make much of a difference in graphics performance.

The Lunar Lake GPU also has a peak clock speed of just 2.05 GHz, down 200 MHz from Meteor Lake's 2.25 GHz and way, way below AMD's potential 2.9 GHz. AMD also has twice as many graphics clusters (CUs) as Intel, but half as many shaders per CU, giving peak theoretical compute of 5.94 TFLOPS. That's 41% more than the 4.20 TFLOPS on Lunar Lake and 29% more than Meteor Lake's 4.61 TFLOPS. Theoretical TFLOPS rarely tell the whole story, so it will be interesting to see how the new Arc Graphics 140V stacks up against the previous Arc Graphics; Intel says the new chips should be universally faster.

All three laptops support DDR5 (or LPDDR5) memory, but the actual memory used is often up to the laptop vendors — though not on Lunar Lake. We're using Asus Zenbook models in all three cases, and both the Meteor Lake and Strix Point models have DDR5 running at 7500 MT/s (the MTL says 7467 MT/s, but close enough). With dual-channel memory that works out to 120 GB/s of raw memory bandwidth — a far cry from what you get with even a modest dedicated GPU, but it should be reasonably sufficient for a laptop with integrated graphics. The new Lunar Lake laptop has 8533 MT/s LPDDR5X memory integrated onto the package, yielding a moderately higher 136.5 GB/s of bandwidth.

As I said, we're looking at apples and oranges in some respects, as there are many differences in the core specifications. Still, it's what we have available, so let's go ahead and run some gaming and graphics benchmarks to see where things land.

We have three laptops, all from Asus, and the MyAsus app also allows you to set a fan speed profile. The default is "Standard," but we've found that selecting "Performance" seems to increase the TDP limit and can help quite a bit with gaming performance. We initially tested each laptop with both the Performance and Standard fan modes (flip through the galleries to see the framegen results), but we'll stick with just Performance mode for the additional non-framegen testing. Windows 11's power profile was set to "Balanced" mode for all the testing shown here.

First up is 3DMark Time Spy. I don't like using synthetic benchmarks for normal GPU testing, but in this case it's a reasonable way to try and remove driver issues from the equation. Every GPU manufacturer should be intimately familiar with 3DMark, and ensuring the drivers work optimally is standard procedure. If the drivers are fully tuned for any games that we test, we'd expect the results to at least echo what Time Spy shows.

And what it shows is that, despite having theoretically less GPU compute available, Lunar Lake's Battlemage GPU comes out on top. It's not massively faster than the other chips, but it shows 13% higher performance than the Radeon 890M when both are using the MyAsus Performance mode fan profile. Compared to the previous generation MTL GPU, it's also 14% faster. Putting all three laptops into the Standard mode fan profile does change things a bit. All three laptops run slower, and the Lunar Lake GPU is now only 5% faster than the 890M but still 12% ahead of the Meteor Lake GPU.

One critical item to point out is that memory bandwidth on Lunar Lake is 13–14 percent higher than on the other two laptops, so instead of showing true gaming potential, 3DMark Time Spy might actually end up scaling more with bandwidth. Some games do behave that way, but plenty of others depend more on the GPU computational abilities rather than bandwidth. So let's check out some games...

Black Myth Wukong is a recent release, which makes it a good showcase for any potential driver concerns. We're using Intel's latest preview drivers for the Lunar Lake system, AMD's 24.8.1 drivers on the Ryzen AI / Strix Point laptop, and the latest publicly available Intel 6077 drivers for Meteor Lake. (6078/5736 drivers just came out today for the retail Lunar Lake launch, but we used 6077 from late last week.)

Starting with the native 720p results, the standings don't look the same as 3DMark. The new Lunar Lake GPU still comes out 11% ahead of the Meteor Lake GPU, but now AMD's 890M ends up 10% ahead of Lunar Lake. This is the issue with synthetic graphics tests like 3DMark versus real-world gaming performance. We'll want to look into testing more games to come to any firm conclusion.

Bumping the resolution to 1080p and enabling quality mode upscaling in both XeSS and FSR3, AMD still comes out on top, with slightly higher XeSS performance compared to FSR3 performance. However, XeSS 1.3 uses larger upscaling factors so that's likely the cause. The two Intel GPUs also show better performance with XeSS than FSR3, though the biggest gap remains less than 10%.

Last in the gallery are the FSR3 results with frame generation enabled, which we retested as there were some anomalies. Things take a turn for the worse on Intel's new GPU here, as framegen does something for the Arc 140V, but it's definitely not working that well. Lunar Lake sees a 17% increase in fps — and half of the framegen frames have no user input so it will feel noticeably worse. That's in stark contrast to Meteor Lake, which improved by 50%, and AMD's 890M that achieved a 63% higher result than the non-framegen 32 fps.

Cyberpunk 2077 has been around for a while, but it just (finally) received the promised FSR3 support, so we'll see how that goes. But let's start with the native 720p result, which removes any upscaling or framegen shenanigans from the equation.

Of our tests, this represents the best results from Intel's new Arc Graphics 140V. It's 32% faster than the previous Arc Graphics solution found in Meteor Lake, and since both are running in basically the same Asus 14-inch chassis, these are directly comparable. AMD's chip still beats Lunar Lake by 18%, but that's at least partially influenced by the larger 16-inch chassis that may have a higher TDP with our test settings.

Bumping to 1080p and enabling upscaling, again noting that XeSS 1.3 uses a larger scaling factor than FSR3 (and DLSS), the Arc 140V takes the top spot with 51 fps, narrowly edging past the GPU chip. AMD's performance is basically the same with XeSS as with FSR3, while Intel (not surprisingly) sees a bigger boost with XeSS. That applies to both Meteor Lake and Lunar Lake, with LNL getting a 6% increase while MTL gets a 9% boost.

Turning on FSR3 framegen completely changes the picture, however, with the Radeon 890M dominating the competition. It's 47% faster than MTL in performance mode and 27% faster in standard mode — so in this case, engaging the performance mode fan profile in the MyAsus app boosts performance by over 40%. Again, chassis size looks to be a factor, with the MTL laptop showing a 24% boost in performance while the newer LNL laptop gets a 14% boost.

As noted with Black Myth Wukong, Lunar Lake doesn't seem to appreciate FSR3 framegen much right now. It saw basically no benefit at all. AMD's GPU meanwhile got a 60% boost to fps in performance mode, and Meteor Lake saw a 54% increase. It's odd that framegen, which just uses GPU shader computation as far as we're aware, doesn't seem to want to run on Lunar Lake right now.

Last, we have Shadow of the Tomb Raider, an older game that should be good and fully optimized for any reasonable drivers. And, interestingly, we get a chart that looks more like 3DMark. There's no framegen this time, and only XeSS (or DLSS if you have an Nvidia GPU), so we opted to include native 1080p performance.

At 720p, the new Lunar Lake GPU claims the top spot with 74 fps, leading AMD's 890M by 6% and beating the Meteor Lake GPU by 25%. That's a good result for what is still likely a power constrained GPU — and having fewer CPU cores probably helps Lunar Lake's GPU make the most of the horsepower it has on tap.

Our second chart is with XeSS Quality mode enabled. Now, I think this is XeSS 1.2 or 1.1, not the latest 1.3, so the scaling factor is 67% — 2.25X upscaling. And the DP4a mode used on the Meteor Lake and AMD GPUs doesn't tend to look as good as the XMX mode running on the Arc 140V. AMD and Meteor Lake end up basically tied, while the new Arc 140V holds a 25% lead.

Finally, we have the native 1080p medium results. Overall, the Core Ultra 9 288V takes top honors with 51 fps in performance mode and 40 fps in standard mode. That's also the biggest improvement from engaging the higher speed fan mode, a 24% increase. The Core Ultra 7 155H only increased by 14%, while the Ryzen AI 9 HX 370 saw a 15% increase. The net result is that Lunar Lake comes out 13% ahead of Strix Point with both performance modes but only 5% ahead in standard mode. It's also 21% faster than Meteor Lake in performance mode and 11% faster in standard mode.

First Impressions of Lunar Lake's Arc Xe2 Graphics

With updated testing results that don't use FSR3 frame generation in Black Myth Wukong and Cyberpunk 2077, our overall impression of Lunar Lake's GPU improves quite a bit. It's not a massive generational performance improvement compared to Meteor Lake, but the 28–30 watt TDP of the two chips we tested probably limits how far Intel can push performance regardless — and memory bandwidth also makes a difference.

There are, however, still some oddities present, things which will probably end up being fixed in Intel's drivers. Why didn't framegen work at all on Lunar Lake, or work very poorly, when it seems to do just fine on Meteor Lake? It doesn't make sense for Intel to intentionally break AMD's tech, but it's also weird that it doesn't seem to work in general. We've tested FSR3 framegen on old Nvidia GTX cards and it tends to work fine, so why would Lunar Lake be any different?

Not shown in the charts, we also ran the same graphics tests on the 'lesser' Core Ultra 7 258V. The main differences between that chip and the Core Ultra 9 288V are the base power (17W vs 30W) and maximum clocks (100 MHz higher for the GPU clock on 288V). In our testing, at least in the standard fan mode, the two chips are within spitting distance of each other on our graphics tests. So, you can safely opt for the slower and less expensive Lunar Lake chip and only give up a very small amount of performance — which makes sense. Lunar Lake doesn't have lots of CPU cores even on the Core Ultra 9 288V.

Overall, the graphics performance from Lunar Lake and the Arc Graphics 140V, aka Arc Xe2 Battlemage, doesn't look particularly amazing, but we didn't expect to see much since we're dealing with integrated graphics. Power and memory bandwidth constraints often eclipse other factors. We've seen claims of double the performance (or at least performance per watt) on various integrated GPU solutions in the past, and those rarely pan out in the real world. With its current focus on ultraportable, low voltage, long battery life laptops, this is hardly the best showcase for Intel's next-generation Battlemage GPU architecture.

We hope that the results for the future desktop cards end up being far more impressive once Battlemage GPUs begin shipping. We also hope that driver issues and oddities continue to get resolved. Intel has improved in that area since the initial Arc launch, but AMD and Nvidia remain far more consistent in both game support and performance. There are times when Lunar Lake might come out ahead of AMD's Radeon 890M, but we suspect that over a larger test suite, Intel will encounter far more frequent anomalies in terms of game support and optimizations.

AMD's flagship Ryzen AI 9 HX 375 CPU was benchmarked in Geekbench 6 with potent results. Initially discovered by HXL on X, the Zen 5 CPU's single-core and multi-core results were high enough to outperform most of Intel's fastest mobile CPUs, featuring the same benchmark results as a desktop Core i5-13600K.

The HX 375 produced a single-core result of 2,864 points and a multi-core result of 15,012 points on an HP OmniBook Ultra Laptop 14 paired with 32GB of DDR5 memory. The Zen 5 chip's results were good enough to outperform Intel's flagship CPUs from its Meteor Lake lineup and almost good enough to surpass Intel's Raptor Lake Refresh mobile flagship.

The Ryzen AI 9 HX 375 was 5% faster than the Core i9-14900HX in Geekbench's single-core benchmark. However, the Core i9-14900HX was 7% faster in the multi-core benchmark. Compared to the Core Ultra 9 185H, the HX 375 lead was much more substantial, 27% faster in the single-core benchmark and 25% faster in the multi-core test.

We don't have official Lunar Lake Geekbench 6 benchmarks yet, but based on some leaked results, the flagship 288V appears to do single-core results in the 2,900 range, which would outperform the HX 375 by a couple of percent. However, multi-core results are generally much worse, in the 10,000 - 11,000 range.

Regardless, the Ryzen AI HX 375 can, at worst, match the performance of Intel's fastest mobile CPUs and, at best, vastly outperform its primary Intel competitors, Lunar Lake and Meteor Lake, specifically in the multi-core results. For a desktop comparison, the HX 375 matches the Core i5-13600K in terms of multi-core performance and the Core i9-14900 in terms of single-core performance.

The Ryzen AI 9 HX 375 is the flagship chip from AMD's Ryzen AI 300 lineup. The chip shares the exact specifications as the HX 370, except for the XDNA2 NPU, which has been improved by 10% and features 55 TOPS of performance rather than 50. Expect to see the Ryzen AI 9 HX 375 inside gaming laptops and mobile workstation devices.

Beelink has launched its latest SER9 mini-PC, powered by AMD's Ryzen AI 300 (Strix Point) processor and priced at a whopping $999. If one takes the official listing page at face value, the SER9 is supposedly discounted from $1,249. When similarly-specced rivals are available for much cheaper, it's understandable why Beelink's price point is somewhat controversial.

While the pricing is questionable, the SER9's features don't seem bad. Using AMD's top-of-the-line Ryzen AI 9 HX 370 processor, the device has 12 cores, four Zen 5 cores, eight Zen 5c cores, and 24 threads. There's also a Radeon 890M iGPU onboard with 16 RDNA 3.5 CUs clocking up to 2,900 MHz. The Ryzen AI 9 HX 370 should reasonably serve you for productivity and casual gaming.

The Ryzen AI 9 HX 370 is accompanied by 32GB of LPDDR5X-7500 memory. According to Beelink, the configuration consists of four memory chips to deliver more performance than a single chip, which is common in some SFF PCs and laptops. However, this RAM is also soldered, so upgrades are out of the question.

Beelink SER9 Mini-PC Specifications

The SER9 comes with a 1TB Crucial M.2 NVMe SSD. Beelink doesn't specify the model, though. Nonetheless, you can replace it with your own. The device has an M.2 2280 slot that runs at PCIe 4.0 x4 and accepts SSDs up to 4TB.

Regarding network connectivity, the SER9 offers Wi-Fi 6 (AX200) and Bluetooth 5.2. Alternatively, if you prefer wired over wireless, you can hook an Ethernet cable up to the 2.5 Gigabit Ethernet port. Display connections on the SER9 include one HDMI port and one DisplayPort 1.4 output.

Another favored feature for some users will be the built-in speakers and microphone array. In scenarios where hooking up these peripherals would be a hassle, especially in an enterprise environment, these integrated audio solutions make a fair bit of sense. However, their inclusion certainly doesn't justify the price point we're seeing here on its own—particularly when some of the cheaper alternatives, like the AOOSTAR Gem10, include OCuLink for yet-improved external GPU performance.

The first review of a mini PC using AMD’s new Ryzen AI 9 HX 370 CPU has arrived, roughly two months ahead of the expected release of the tiny desktop computer. It’s the Soyo S9 AI mini PC, one of several such products in the pipeline for release. YouTube creator ETA Prime got his hands on a prototype of the computer and put it through its paces.

The Ryzen AI 9 HX 370 Strix Point CPU in the Soyo S9 AI uses AMD’s new Zen 5 architecture, with a total of 12 cores. Four of those are Zen 5 cores, while eight are slightly denser Zen 5c cores. Testing confirms the new CPU has 16 Compute Units (up from 12), which helps with performance.

The Soyo S9 AI mini PC includes two USB-A ports and a single USB-C port on the front. At the time of review, however, the front USB-C port was not functioning. On the rear of the computer are two HDMI ports, an Ethernet port, two more USB-A ports, and a USB-C port. ETA Prime could not verify if the rear USB-C port would support the USB 4 standard, but was told the front port should (when it's made functional.)

Among other positives for the upcoming mini PC, according to ETA Prime, the computer has a total of three M.2 NVMe PCIe 4 SSD slots. One of those wasn’t working on his review prototype, but that’s an issue sure to be resolved before production. 

The computer utilizes LPDDR5X RAM and is available with 16GB, 32GB, or 64GB. The downside here is that upgrading the RAM won’t be a possibility, since the LPDDR5X is soldered to the mainboard.

It’s important to note that this mini PC lacks a discrete GPU, but that doesn’t mean you should discount its capabilities. Graphics are upgraded from RDNA3 to RDNA 3.5, and ETA Prime found several AAA games played with impressive frame rates at 1080p resolution. The reviewer tested gaming titles like Cyberpunk 2077, Ghost of Tsushima, Horizon Forbidden West, and more. In Prime’s words, “With the new Radeon 890M iGPU this mini PC can run AAA Games like a BOSS.”

*data in table compiled by VideoCardz

Granted, he couldn’t achieve those frame rates with the highest graphics settings, but neither was he forced to dial everything down to the bare minimums, either. The last-generation series, Prime pointed out, needed either lower-quality FSR profiles or reduced resolution.

The exact release date and pricing for the Soyo S9 AI mini PC remain unannounced, but the computer is anticipated to drop by mid-October. Soyo is apparently on the fence about pricing and asked Prime to poll his viewers for their thoughts on what they’d be willing to pay for such a PC.

EIDT 8/25/2024 11:50am PT: Added correct attribution. 

Earlier today, the product page for the upcoming GPD Pocket 4 handheld/tablet/mini laptop went live on GPD's website. The product page reveals a full set of system specifications and a fairly expansive list of optional expansions and potential use cases, focusing on creativity and productivity.

While the gaming utility is somewhat understated, the Pocket 4's status as the first Ryzen AI 9 HX 370 handheld, physical keyboard, multi-touch interfaces, and motion controls should all have it fairly equipped for those scenarios, too. Steam Input used with a device with this power and this range of utility shows some promise despite gaming not being the focus of the form factor.

Of the four expansion modules listed on the original product page to be slotted in with the module system, only the microSD card Reader Module is actually included with the device. The other three listed Modules (EIA RS-232 Module intended for use in Terminal Display use cases; KVM Control Module intended for use in data center/server and multi-PC use cases; 4G LTE Module self-explanatory) are all separate purchases.

GPD Pocket 4 Specifications

 *Both these USB-C ports support USB power delivery.

At this point, all that's missing is an OCuLink eGPU Module, though GPD touts the Pocket 4's support for USB4 eGPU docks as an intended use case. However, OcuLink is still faster, so if these Expansion Modules use enough PCIe lanes, there could be more potential.

While the general CPU and iGPU specs are quite impressive, one important caveat is worth noting. This PC can only run its Ryzen AI 9 HX 370 at up to 28 Watts, the default TDP of a CPU that can otherwise be configured to use a 54-watt TDP.

Such a TDP tweak on this handheld is intended. However, a higher TDP could result in dramatically fast battery depletion or overheating, which isn't ideal for its intended handheld enthusiasts or tech professional environments. It also means that the true GTX 1070-esque highs of the new top dog iGPU, the Radeon 890M, aren't quite explorable on this device.

That said, the most impressive aspect of the GPD Pocket 4 was never its hardware power— the sheer versatility of input and use case potential is staggering. All that on top of an 8.8-inch multi-touch tablet display with 97% DCI-P3 coverage and ICC calibration support, and the Pocket 4 looks to be an enticing offer— but no word is yet available on pricing or release date, and the usage of 45W TDP Ryzen AI 9 HX 370 benchmarks on the Pocket 4's page is misleading of GPD.

It's rare that my lasting impression about a laptop is all about style, but here we are. The Asus Zenbook S16 ($1,399.99 to start, $1,699.99 as tested) is a vehicle for the AMD Ryzen AI 9 HX 370 and one of the first AMD Ryzen AI 300 series laptops on the market, but it's also just a nicely designed laptop. Design and performance are both important parts of our considerations for the list of the best ultrabooks.

Asus's new Zenbook uses some proprietary materials, including an aluminum/ceramic mix on the lid that looks straight-up classy. The light gray paired with white accents is pretty, but not overwhelming, and the fact that it houses strong performance shows that beauty isn't just on the outside.

That’s not to say this is a perfect laptop. While it did well on our battery test, Arm-based machines often lasted much longer. And the OLED display, while colorful, isn't as bright as some competitors.

Design of the Asus Zenbook S16

Asus is keeping it simple with the Zenbook S16, to great effect. This is a pretty laptop that puts almost all of the focus on the 16-inch screen. There's flair, but it's used sparingly.

The company is using a material it refers to as "ceraluminum," a combination of ceramic and aluminum. It seems similar to the "plasma ceramic aluminum" we saw in the lid of the Asus Zenbook S13 OLED last year. It makes the lid look stunning, especially up close. The clean lines of the Zenbook logo cut through the "zumaia gray" material, which really looks like it could be pottery up close.

Opening the lid, however, you realize the rest of the chassis feels like a more standard aluminum. The cooling vents are extremely detailed; it's a grid of rounded squares, each with two holes for airflow. In contrast, the rest of the machine is plain, with a silver chassis and white keyboard (with white backlighting), as well as a large touchpad.

The left side of the laptop features two USB 4 Type-C ports, an HDMI output and a 3.5 mm headphone jack. On the right, there's a USB 3.2 Gen 2 Type-A port (a surprise on a notebook this thin) and a full-sized SD card slot.

The Zenbook S16 is 13.92 x 9.57 x 0.47 inches and weighs 3.31 pounds. The HP OmniBook X, a 14-inch laptop with a Qualcomm Snapdragon X Elite processor, is 12.32 x 8.8 x 0.57 inches and 2.97 pounds, making it lighter but thicker. The HP Spectre x360 14, a 2-in-1 running an Intel Core Ultra processor, is 12.35 x 8.68 x 0.67 inches and 3.19 pounds, which is thicker and closer in weight.

On the Mac side, Apple's 15-inch MacBook Air with M3 is 13.4 x 9.35 x 0.45 inches and 3.3 pounds, slightly thinner and about the same weight. The 16-inch MacBook Pro with M3 Max is 14.01 x 9.77 x 0.66 inches and 4.8 pounds, the heaviest of the bunch.

Asus Zenbook S16 Specifications

Productivity Performance on the Asus Zenbook S16

This review happens to be our first look at AMD's new Ryzen AI 9 HX 370, one of the company's very lengthily-named processors in its "Strix Point" lineup. This chip has four Zen 5 cores and eight Zen 5c cores (for a total of 12 cores). It has a base clock of 2 GHz and a max boost of up to 5.1 GHz. As chip companies attempt to one-up each other on AI, AMD is packing 50 TOPS in its NPU, surpassing the 45 TOPS in the Qualcomm Snapdragon X Elite and X Plus (the chip as a whole goes up to 80 TOPS).

In our Zenbook S16 review unit, the Ryzen AI 9 HX 370 is paired with 32GB of RAM and a 1TB M.2 PCIe NVMe SSD.

On Geekbench 6, the Zenbook S16 achieved a single-core score of 2,765 and a multi-core score of 13,282. This was enough to beat the M3 in Apple's MacBook Air, the 16-core Intel Core Ultra 7 155H in the HP Spectre x360 14, and the Qualcomm Snapdragon X Elite X1E-78-100 in the HP OmniBook X. The newest Ryzen fell only to the Apple M3 Max, with scores of 3,200 and 21,711 in the 16-inch MacBook Pro (and, to be fair, a review unit stocked with 128GB of RAM, making the MacBook Air a more apt comparison from a price perspective. While it's unclear that Apple has ever released an official number, various sites have tested the M3 Max to be using significantly more power than the Ryzen AI CPU here, which is rated at 28W and configured to 17W TDP out of the box).

The Zenbook S16 copied roughly 25GB of files at a rate of 908.45 MBps. While that was faster than the HP OmniBook X, every other competitor was faster on this task.

On Handbrake, the Zenbook transcoded a 4K video to 1080p in 5 minutes and 9 seconds. The next closest was the Snapdragon X Elite-based Omnibook X at 5:46. The M3 Max won out at 2:34, but again, that's in a far more expensive machine.

To stress the processor, we run Cinebench 2024 for 10 runs. During the test, the system started at 821.18 points and dipped into the high 700's, and then peaked up to 844.42 points before dipping again into the mid-to-high 700's. The Zen 5 cores ran at an average of 2.67 GHz, while the Zen 5C cores reached 1.9 GHz. The CPU ran at an average of 73.44 Celsius.

AI on the AMD Ryzen AI 9 HX 370

Despite all of the hubbub around artificial intelligence, it's still only useful in so many consumer products. On Windows laptops, applications are still slowly adopting the technologies. While NPUs are often being used for background blur in video conferencing apps, generative AI is often using GPUs (at least, when it isn't looking to the cloud instead).

In a reviewer's guide sent to the press, AMD suggested a number of applications to try that make use of local AI. Chief among them was a beta of Amuse 2.0, which includes image generation with Stable Diffusion, as well as paint-to-image features that are very similar to image creator in Paint on Copilot Plus PCs (though that uses DALL-E). In a way, this app, made to work well with AMD's hardware, can fill in the blanks until Microsoft adds Copilot Plus features to Ryzen AI laptops.

I loaded Stable Diffusion 1.5 into Amuse to test image generation the same way we tested it on Meteor Lake and Snapdragon Elite X laptops. Our go-to prompt, "cat driving a car," took 9.9 seconds. Notably, the software utilized the GPU, rather than AMD's new NPU. (Targeting the CPU, which we also tried, led to worse performance).

That makes it faster than the Intel Core Ultra laptops we tested, which used the GPU (approximately 13 seconds), but slower than a a pair of Surface notebooks utilizing the latest from Qualcomm, which targeted the NPU (approximately 7 seconds).

I also tried a preview of Camo Studio, which uses the NPU to blur backgrounds, though you can already do that with Windows Studio Effects. There are other uses, like AI upscaling in Gigapixel AI, but these are mostly tools for creators. The average person, at the moment, doesn't seem to have a lot of reasons to use local AI features, and should probably focus more on productivity performance and battery life.

Display on the Asus Zenbook S16

The screen on the Zenbook 16 makes a colorful first impression. The 16-inch, 2880 x 1800 OLED touch panel was vivid to my eye, but not as bright as some others. It also goes up to 120 Hz refresh rate.

I used the Zenbook 16 to check out the trailer for Captain America: Brave New World, which is heavy on explosions. They all appeared a fiery orange. In a scene that appears to be in a laboratory, new Cap Sam Wilson uses the shield to fend off two attackers with electric batons, which showed off a bright blue against neon red lights.

Asus's 16-inch panel covers 79.7% of the color gamut (113% sRGB), in range of both the MacBook Air and the HP OmniBook X. The 16-inch MacBook Air's mini-LED display and HP Spectre x360's OLED screen were both more vivid.

At 357 nits, I found the Zenbook 16 to be usable, but competitors do better. The 16-inch MacBook Pro was the brightest at 563 nits, and both the MacBook Air (476.4 nits) and HP Spectre x360 366.2 nits were also brighter.

Keyboard and Touchpad on the Asus Zenbook S16

Despite just 1.1 mm travel, the Zenbook S16’s keyboard is snappy. Don't get me wrong, it's still a membrane keyboard, but it's relatively comfortable and even has a bit of bounce. On monkeytype, I hit 112 words per minute with a 2% error rate, which isn't my best, but is pretty close.

The keyboard is set back slightly further than I'd like because the touchpad is so tall. I got used to it but some of my coworkers found it a bit uncomfortable.

The 5.9 x 3.9-inch touchpad is ample, taking up much of the space between the keyboard and the edge of the deck. While I'm increasingly fond of haptic touchpads, Asus is using a standard physical click here, which feels decent enough. 

All of the typical Windows gestures work well enough, and Asus threw in some of its own. Swiping up and down alongside the right side of the touchpad adjusts screen brightness, while the left side adjusts volume. If you swipe along the top while a video plays, you can scrub the timeline in 5 second increments to move forward or backward. I initially worried that these would get in the way of my using the device, but it wasn't an issue. Those controls are on the very edge, and since I usually start towards the center, I never set them off by accident.

Asus also includes its Asus Pen 2.0 — a stylus with 4,096 degrees of pressure sensitivity — in the box with the laptop. It's a nice add-on, I suppose. But it’s an odd fit for this clamshell laptop, particularly considering the display doesn't fold flat.

Audio on the Asus Zenbook S16

The Zenbook S16 trades volume for detail. Don't get me wrong, it's loud enough, especially if you're sitting right in front of it, but it didn't quite fill an entire meeting room with sound. But when I was listening to Charli XCX's "Von dutch," I was surprised to hear a bit of bass. The speakers, tuned with Dolby Atmos, sounded great when I switched from "balanced" to "detailed," with clear vocals, pounding drums and whirring synths.

Upgradeability of the Asus Zenbook S16

There isn't much to upgrade or repair on the Zenbook S16.

There are 12 screws on the underside of the laptop. Four of them — the ones closest to the palm wrest and the one in the center — are shorter than the rest, but only very slightly so. Be sure to keep your screws organized. The clips were tight, but with patience and a tool, we pried them open.

Once you're in, there are two parts you could replace. The first is the battery, which takes up most of the room inside the system. The other is the SSD. Despite the size of the laptop, there isn't a secondary M.2 slot, so you'll have to replace the drive if you want more storage space. The wireless module and RAM are soldered to the motherboard.

Battery Life on the Asus Zenbook S16

While the Zenbook S16 will likely last you an entire workday, the new Asus device didn't last as long on a charge as competitors based on Arm platforms.

The AMD-based Zenbook ran for 11 hours and 35 minutes on our battery test, which involves web browsing, OpenGL tests, and video streaming over Wi-Fi with the screen set to 150 nits. Only the Spectre x360, with Intel Core Ultra, did worse at 11:01 (and that's not far behind, especially considering that laptop has a smaller battery than the Zenbook).

But the MacBook Air M3 (15:03), HP OmniBook X on a Snapdragon X Elite (16:18), and the 16-inch MacBook Pro lasted an incredible 17:11.

Heat on the Asus Zenbook S16

During our ten runs of Cinebench 2024, we took skin temperature measurements to see how the Zenbook performs under an intense workload.

The keyboard hit a warm (but not uncomfortable) 99.5 degrees Fahrenheit, while the touchpad stayed cool at 87.4 F. The hottest point on the top on the grille at 110 F, but that's not usually a place you put your hands. On the bottom, the hottest touchpoint, near the ventilation, was at 116.2 F.

Webcam on the Asus Zenbook S16

There's a 1080p webcam on the Zenbook S16's top bezel. While it's color-accurate, catching the exact shade of green in my shirt and the blue in my mesh office chair, some details, like my hair, came across a bit fuzzy.

The camera supports infrared to log in with Windows Hello for facial recognition. It's the sole biometric option on the device.

Software and Warranty on the Asus Zenbook S16

Asus includes four pieces of software on the Zenbook S16. The first is MyAsus, which allows you to manage your warranty, and also shows system settings and lets you check for updates. This app now prompts you for an account, which it didn't use to do (and in my opinion, shouldn't need to). Despite its appearance, you can skip this prompt.

ScreenXpert manages your windows across displays, including monitors plugged into your PC. It also has other functions, like letting you turn your camera and microphone on and off. GlideX lets you use a phone or tablet as an additional screen or mirror your phone's screen on the laptop. However, you'll need to pay a subscription fee to avoid ads and get access to the best streaming quality.

StoryCube is a new one. Asus describes it as an "AI media hub" designed to sort your files with a focus on photos and video. It can categorize photos by topic, location, and people. My smartphone has been doing this for awhile, so this isn't exactly groundbreaking.

Otherwise, Asus just has some Windows Store apps pinned to the Start menu, like Spotify, which you'll have to delete if you don't want them.

Asus sells the Zenbook S16 with a one-year warranty. A card included in the box offers one free year of accidental damage protection in exchange for registering the laptop.

Asus Zenbook S16 Configurations

We tested the Asus Zenbook S16 with an AMD Ryzen AI 9 HX 370, 32GB of RAM, and a 1TB SSD, which goes for $1,699.99.

A cheaper version sold at Best Buy is $1,399.99 with a Ryzen AI 9 HX 365 and 24GB of memory, as well as a white chassis.

Bottom Line

The Asus Zenbook S16 is a solid showcase for AMD's new AI-focused processors, even if AI applications themselves aren't quite ready to excite most people.

The AMD Ryzen AI 9 HX 370 was competitive in our productivity benchmarks, and the laptop's ceramic/aluminum hybrid lid is still unique and interesting, without being garish. (It's a fine line to walk.) The roomy touchpad is nice, and it's cool to see Asus manage to fit a USB Type-A port in a laptop this thin.

It's hard to say that 11 and a half hours of battery life is a downside, but it does show that x86 laptops are lagging a bit in that department behind the likes of Apple and Qualcomm, both using Arm-based designs. Over 11 hours is great; over 15 hours is better.  Additionally, while the OLED display is extremely colorful, it's not as bright as some competitors.

But if you want a 16-inch PC and are looking for an AMD machine with the NPU with the most TOPS out there (so far), the Zenbook S16 with Ryzen AI 9 is a strong choice. That being said, there are only so many applications that use those TOPS, and only so many that directly target the NPU. If you're an AI enthusiast, you'll be ready. For everyone else, the Zenbook S16 is still a pretty nice laptop.

MORE: How to Buy a Gaming Laptop

MORE: Best Gaming PCs

MORE: Best Ultrabooks and Premium Laptops

More details about AMD's Ryzen AI "Strix Point" CPUs have been coming to light over the past few weeks, and today we got official confirmation of a third Ryzen AI 9 300 Series CPU, the Ryzen AI 9 HX 375, to join the Ryzen AI 9 HX 370 and Ryzen AI 9 365. These mobile Strix Point CPUs blend Zen 5 and Zen 5C cores (the exact ratio varies depending on the chip) with onboard XDNA 2 NPUs for AI workloads and an optimized RDNA 3.5 architecture for the new iGPUs. 

A recent Zen 5 showcase showed off benchmarks of the Ryzen AI 9 HX 370, outperforming its contemporaries by a fair margin. While manufacturer-supplied benchmarks should always be taken with a grain of salt, other recent Ryzen AI 9 HX 370 benchmarks lend credence to AMD's official numbers. Since the new Ryzen AI 9 HX 375 slots in directly above the Ryzen AI 9 HX 370, this bodes well for the near future of AMD-powered laptops and handheld PCs, though we don't anticipate a dramatic performance difference between HX 375 and HX 370.

Confirmed AMD Ryzen AI 9 CPU Specifications So Far

The main difference between the Ryzen AI 9 HX 375 and its former highest-end counterpart, the Ryzen AI 9 HX 370, is a slightly-amped up NPU with 5 TOPS more throughput to throw toward AI tasks.

The Ryzen 9 HX 370 does already seem to be quite competitive with the likes of the Intel Core Ultra 9 185H, after all, and even outpaces the last-gen Ryzen 9 7945HX3D in synthetic single-core performance. 

Ultimately, we'll have to wait and see how these Ryzen AI 300 Series CPUs are used across different devices and how the real-world pricing difference between HX 370 and HX 375 manifests to determine whether this ends up being a harmless extra SKU or an attempt to nickel the end customer for five extra TOPS of NPU throughput. For now, though, the existing benchmarks for these new chips look promising for AMD, and another high-end laptop CPU is always nice.

AMD's 12-core Zen 5 mobile flagship, the Ryzen AI 9 HX 370, has debuted on PassMark. The Zen 5-based processor showed competitive performance against Apple's top-of-the-line M3 Max CPU, beating the 14-core variant and approaching the 16-core in the single and multi-threaded scores. The chip surpassed Intel's outgoing Raptor Lake high-performance HX series CPUs, beating the Core i7-14700HX. All comparisons were taken from the multi-core results in PassMark's list of CPU benchmarks.

The Ryzen AI 9 HX 370 scored 37,699 points in PassMark's multi-threaded benchmark and 4,213 points in PassMark's single-threaded benchmark. The Zen 5 CPU's score barely surpasses Apple's 14-core M3 Max, beating it by 4%. Apple's beefier 16-core M3 Max manages to get ahead of the Ryzen AI 9 HX 370, but only by 7%, making the Zen 5 CPU extremely competitive with both parts.

The Ryzen AI 9 part was also very performant against some of Intel's outgoing high-performance Raptor Lake parts. Despite being optimized for thin-and-light use — and having fewer cores, the Zen 5 chip outperformed Intel's 20-core Core i7-14700HX by 4%.

The only chips the Ryzen AI 9 chip struggled to compete with were Intel's flagship Core i9 HX-series CPUs and AMD's high-performance Dragon Range Ryzen 9 parts. For instance, Intel's Core i9-14900HX outperformed the Ryzen AI 9 HX 370 by a healthy 23% and AMD's Ryzen 9 7940HX by an even greater 42%.

AMD's Ryzen AI 9 HX 370 is its latest flagship mobile part, featuring its new Zen 5 CPU architecture. The chip comes with two core clusters consisting of four Zen 5 cores and eight compact Zen 5c cores. Clock speed maxes out at 5.1 GHz boost, with a TDP of 28W (but configurable from 15W to 54W). The chip also has 36MB of cache and AMD's top-of-the-line Radeon 890M integrated graphics with 16 CUs running on the more power-efficient RDNA 3.5 architecture.

PassMark's performance numbers back up AMD's claims that the HX 370 is going toe-to-toe with Apple's M3 silicon. If PassMark's results indicate real-world performance, AMD will finally have a chip that can match Apple's silicon performance in a power envelope similar to that of Apple's M3-powered Macbooks.

AMD’s next-generation mobile and desktop chips are expected to launch in July, and we’ve seen some benchmark results for the top-end Ryzen AI 9 HX 370 gradually appearing online. Hardware detective HXL shared a photograph of a Cinebench R23 test showing AMD’s latest processor running Windows 10. As with all leaks, treat the results with a pinch of salt. The Ryzen AI 9 HX 370 reportedly achieved a single-core score of 2,010 points, while its multi-core performance hit 23,302.

The HX 370 seemingly beats both the AMD Ryzen 9 7945HX3D and the Intel Core Ultra 9 185H in single-core performance, but it’s still outclassed by the former in terms of multi-core score owing to the Ryzen 9 7945HX3D’s eight extra threads. It’s also interesting to see the new AMD chip getting a higher single-core score than the Apple M3 Max and a multi-score performance within a stone’s throw of Apple’s top laptop chip.

We do have an issue with the conducted test, though, as AMD has dropped support for Windows 10 on its Ryzen AI 300 series processors. Nevertheless, resourceful PC enthusiasts could find a way to make it work, or maybe AMD did not include Windows 10 in its documentation, but the hardware could still run it.

Cinebench R23 focuses more on productivity; thus, these results show that the AMD Ryzen AI 9 HX 370 packs some muscle. However, the leaker didn’t indicate what other parts this system has, so we don’t have the complete picture of how it would perform in the real world.

With the launch just a few weeks away, many AMD fans would be excited to get their hands on its latest processors. But we recommend holding off on your purchases until tech reviewers get their hands on these laptops and run them through their paces. That way, you know which options to look at and can get your money’s worth on the new computer you’ll buy.

AMD is expected to launch the Ryzen AI 300 (Strix Point) CPUs on July 28, and reviewers are likely to conduct their final benchmark tests on laptops sporting these chips. One test, shared by hardware detective HXL, showed results for the Ryzen AI 9 HX 370 installed on an Asus ProArt P16. The tested laptop runs Windows 11 Home on the Standard power plan with 32GB of RAM.

The Geekbench 6 test returned a 2,879 single-core and 14,888 multi-core score, giving it a higher single-core score than the previous generation AMD Ryzen 9 7945HX3D and putting it close to the Intel Core i7-13700KF. However, its multi-core score is still almost 2,000 points behind the former and 3,000 points behind the latter. Nevertheless, this is the highest result, so far, that the chip clocked on Geekbench compared to previous test runs.

Here’s how the AMD Ryzen AI 9 HX 370 compares with comparable chips. We will also include Geekbench leaks for Intel’s upcoming Lunar Lake chips to compare them. Treat the results with a pinch of salt since neither Strix Point nor Lunar Lake has launched, and these are leaked benchmarks.

The AMD Ryzen AI 9 HX 370 beats all the other chips in terms of single-core score, but it’s still outclassed by the previous-generation Ryzen 9 7945HX3D in terms of multi-core performance, owing to the four extra cores and eight extra threads on the older chip. Nevertheless, the defining feature of the Ryzen AI 9 HX 370 is its built-in NPU, which allows it to reach up to 50 TOPS.

However, even if it has more than the minimum requirement to get Microsoft’s Copilot+ PC branding, it won’t get those features at launch. Nevertheless, AMD confirmed that Copilot+ features will arrive later this year via a free Windows Update.

These numbers seem promising for AMD’s latest generation of laptop processors. However, these numbers mean nothing without actual real-world results. So, it’s always wise to wait for theoretical and practical benchmarks from reviewers after the laptop has hit store shelves. That way, you’re sure that the devices being tested are randomly picked from the retail stack and that you’re not looking at the results from cherry-picked devices that won the silicon lottery.

The Ryzen AI 9 HX 370 12-core mobile CPU was spotted in yet another leaked benchmark run, combined with a new Ryzen 7 part dubbed the Ryzen AI 7 Pro 360. As reported by IT Home, the two processors were spotted in the CrossMark benchmark, where the Ryzen 9 AI HX 370 outperformed Intel's outgoing Core Ultra 9 185H processor. However, the Ryzen AI 7 Pro 360 showed underwhelming results, beaten by a two-generation old Ryzen 9 7940HS with the exact core count.

Technically, we have already reported on the Ryzen AI 7 Pro 360, but in our previous coverage, it was known as the Ryzen AI 7 Pro 160. If you don't know by now, the 100 Series nomenclature was originally going to be the name of AMD's first generation Ryzen AI lineup, but AMD switched the numbering scheme from 100 to 300 at the last minute. As a result, the Ryzen AI 7 Pro 160 and the Ryzen AI 7 Pro 360 have the same chip.

We now have two confirmations that the Ryzen AI 7 Pro 360 is an 8-core Zen 5 CPU, boasting two core clusters featuring odd-core counts. One cluster has three regular Zen 5 cores, while the other sports five smaller Zen 5c cores. The 8-core configuration is backed up by just 8MB of L3 cache and 8MB of L2 cache (supposedly), and a Radeon 870M GPU, which is suspected of coming with 8 CUs (half the core count of the Radeon 890M).

The posted PassMark benchmark results compared four CPUs: the Ryzen AI 9 HX 370, Ryzen AI 7 Pro 360, Core Ultra 9 185H, and Ryzen 9 7940HS (mistakenly labeled as a Ryzen 7 SKU on the image).

The Ryzen AI 9 HX 370 scored 1,713 points overall, featuring 1,549 points in the productivity segment, 2,128 points in the creativity segment, and 1,212 points in the responsiveness segment.

The next fastest chip in the group was the Meteor Lake-based Core Ultra 9 185H, sporting 16 cores and 22 threads. The Intel chip scored 1,693 points overall, 1,549 points in the productivity segment, 1,981 points in the creativity segment, and 1,391 points in the responsiveness segment.

The third fastest chip was AMD's previous generation, Ryzen 9 7940HS, sporting eight Zen 4 cores. Overall, the chip scored 1,623 points, 1,561 points in the productivity segment of the CrossMark benchmark, 1,792 points in the creativity segment, and 1,369 points in the responsiveness segment.

The Ryzen AI 7 Pro 360 is the back of the pack, the slowest of the four CPUs in this benchmark run. The Zen 5 mobile processor scored 1,373 points overall, 1,267 points in productivity, 1,639 points in creativity, and 1,023 points in responsiveness.

The Ryzen AI 9 HX 370's results are not surprising. The chip outperformed all the other CPUs in performance thanks to its twelve Zen 5/Zen 5c cores. Intel's Meteor Lake Core Ultra 9 185H, despite its larger core count, should be considered a previous-generation product now. Intel is getting ready to release its Ryzen AI 300 series competitor, Lunar Lake, later this year, hopefully closing the gap with AMD and its speedier Zen 5 architecture.

The Ryzen AI 7 Pro 360's results are more surprising. Previously, we saw the chip outperform the Ryzen 9 8945HS (a newer variant of the 7940HS) in Geekbench. The story is much different here, with the pro series Zen 5 chip noticeably underperforming compared to the Ryzen 9 7940HS. The 7940HS is 18% faster compared to the Ryzen AI 7 Pro 360 in the overall benchmark score.

This shows that we can't take these results too seriously; the Ryzen AI Pro 360 in this test could potentially be an engineering sample with much lower clocks than its future production counterpart. The performance results aren't surprising, given that the Pro 360 sports half the L3 cache capacity compared to the Ryzen 7 9740HS. If CrossMark is an L3-sensitive benchmark, that would explain the performance deficit.

Asus confirms via Weibo that the company will showcase its next-generation notebooks using AMD's Ryzen AI 300 processors on July 17. Originally scheduled for July 8, the event will provide an exclusive to showcase these laptops throughout its range. While the reasons for the delay have not been mentioned, it is likely because AMD has allegedly delayed the Ryzen AI 300 launch to July 28, a day before the Ryzen 9000 processors. Retail giants like Best Buy also moved their availability dates accordingly. 

Some known models are from the TUF Gaming, ROG Zephyrus, Zenbook, and ProArt series in different configurations and builds. What's known so far via the listing on Asus' website is that the Ryzen AI 9 HX 370 and the Ryzen AI 9 365 are used in these models. Specific models like the ProArt PX13 series, either of the two with variable storage and RAM options, while some, like the ProArt P16, are paired with the Ryzen AI 9 HX 370. These models also feature a dedicated Copilot Key to function with Windows 11's AI integration.

The ROG Zephyrus G16 GA605 has four variations using the identical AMD Ryzen AI 9 HX 370 processors and GeForce RTX 4060. However, its comparison at the time of writing does not show the difference between all four variants. The TUF Gaming A16 FA608 has two variations with the same Ryzen AI HX 370 CPU but between RTX 4070 and RTX 4060 notebook graphics and the choice between two memory and storage options. The TUF Gaming A14 FA401, which lacks a pre-installed operating system, comes with an RTX 4060 or RTX 4050 paired with the Ryzen AI 9 HX 370.

The Vivobook S16M5606WA is one of the Vivobook models with the Ryzen AI 9 APU. It provides options with no OS and pre-installed Windows 11 Pro for Business and Pro, utilizing the onboard AMD Radeon graphics from the Ryzen AI 9 HX 370.

Speaking about the ProArt series marketed towards content creators, the PX13 HN7306 has a 13.3-inch 3K OLED monitor with a rotating hinge, enabling users to use it as a tablet since it has a touch screen with stylus support. There are three variations of the PX13, each having a choice between multiple notebook graphics from AMD and Nvidia with variable storage and RAM options.

The ProArt P16 H7606 has a 3840 x 2400 16:10 OLED display with up to 60 Hz refresh rate but does have a touch screen with stylus support. It does not have a rotating hinge, and graphics options are limited to the RTX 4070 and Radeon 890M graphics. Based on the comparisons, the P16 has triple fan designs to allow up to 120W CPU TDP and GPU TGP, with the PX13's cooling setup rated to 115W. Both notebooks are advertised to use Thermal Grizzly liquid metal and built-in dust filters.

After AMD's AI notebook chips are announced, many notebook makers will inevitably showcase upcoming models using these chips. Meanwhile, Intel should be shipping Lunar Lake by Q3 and likely have its partners showcasing and retailing by January, assuming no delay. Naturally, one would like to see how these processors perform compared to notebook processors without the neural chips and the Snapdragon X Elite. 

The Ryzen AI 300 series still has a few weeks until its release, but new Geekbench scores bode well for the chips. Two new engineering samples of the Ryzen AI 9 HX 370 APU were found on Geekbench today, reaching scores beating the highest-end Zen 4 mobile processors, and running at their full speed.

The last time Strix Point (AMD's codename for the upcoming Ryzen AI 300 series) was seen in Geekbench leaks, it turned in impressive test scores but was running in Silent Mode. This power-saving meant the chip only reached a core clock of 3.67 GHz, well shy of its recorded boost clock of 4.2 GHz. Today's test scores let the processor fully loose, with two chips hitting 4.5 and 4.8 GHz. The 4.5 GHz chip still holds the outdated "Ryzen AI 9 HX 170" name for the 370, while the other is an unnamed engineering sample. 

As in our previous coverage, below is a table of the Ryzen AI 9 HX 370's best result in this latest batch of results (the 4.5 GHz test) compared with the June 7th results, as well as its predecessor, the Ryzen 9 8945HS, and AMD's fastest available mobile processor, the Ryzen 9 7945HX3D. 

In just the two weeks between the Silent Mode benchmark and today's results, Strix Point has seen some improvement. Turning off Silent Mode and likely some level of driver optimization and development contributed to an 11% jump in single-core scores and a 4% gain in multi-core. With an expected due date of July 15th, AMD and its OEM partners are likely to continue optimizing performance until better results are unlocked alongside the retail launch of the 300 series. 

As expected from a generational leap, the HX 370 soundly beats the 8945HS, giving a 19% increase in single-core performance and a 25% boost in multi-core. This is more in line with the 16% improvement in IPC between Zen 4 and Zen 5 CPU architecture that AMD has been advertising. 

More impressively, the Ryzen AI 9 HX 370's single-core results squeezed past the Ryzen 9 7945HX3D, beating its 2,820 score with a 2,833. This is a bit of a shock, considering the 7945HX3D contains AMD's 3D V-Cache technology, which powers AMD's best CPUs for desktops ( for more on 3D V-Cache, see our explainer) and Dragon Range's higher TDP over Strix Point. What's unsurprising is the 7945HX3D's 11% lead in multi-core testing, thanks to its four additional cores. 

Geekbench GPU results also popped up today for the integrated graphics within the Ryzen AI 9 HX 370 APU, though the Radeon 880M is named rather than the 890M which is confirmed to be in the HX 370. Results in Geekbench 6's OpenCL and Vulkan tests are in line with what we reported last week on the Radeon chip's early TimeSpy results. With a Vulkan test score of 33,849, the "880M" falls in line with the performance of discrete GPUs like the GTX 1650 Max-Q (34,169) and beats Intel's desktop Arc A380 (31,766). 

We would assume that the integrated graphics were listed incorrectly in the Geekbench tests, as the HX 370 has the Radeon 890M inside, but engineering sample weirdness or some other problem could be to blame. The RDNA 3.5-based Radeon 890M GPU has 16 compute units, which allows it to soundly beat its predecessor, the Radeon 780M. 

AMD's Strix Point gets better every time we see it. Even with a confusing last-minute name change from the 100-series to 300-series to have a bigger number than Intel, the benchmark results it turns in are good enough to forgive AMD's clout chasing. The Ryzen AI 9 HX 370 will reportedly be released in laptops starting July 15th. And if launch performance is anything like what we're seeing today, it should be a big win for AMD over Intel in the laptop market in both graphics and processing power.

Laptops featuring AMD's Ryzen AI 300 Series "Strix Point" CPUs will launch on July 15, two weeks ahead of a July 31 launch date for the Ryzen 9000 series "Granite Ridge" CPUs. That's according to leaks and information from several sources indicate, including laptop listings from Best Buy found by Twitter by user @9550pro, and WCCFTech found Ryzen 9000 series pricing Canada Computers and a Philipine site; B&H Photo also has a Ryzen 9 9950X listing "coming soon."

In addition to these retail listings, AMD updated its 3D V-Cache optimizer driver last week. That's linked to Ryzen 9000X3D parts, which seems set to further iterate on past X3D CPUs and their boosted gaming performance, but we don't expect the X3D variants to arrive until at least a few months after the initial Zen 5 parts. July 31 will seemingly be the launch date for the Ryzen 9000 non-X3D processors.

Pricing details stem from non-U.S. listings, so we have to convert those to USD to get estimated MSRPs. The least expensive parts appear to start at around $300 USD for the Ryzen 5 9600X, and range up to $600 USD for the Ryzen 9 9950X. There's probably about $20 leeway, up or down, for the U.S. retail launch. Note that the numbers come from Canada Computers and Bermor Techzone in the Philipines, while B&H doesn't yet show a price.

The Ryzen AI 300 CPUs will of course only be sold in laptops, but the Asus Zenbook S 16 listing on Best Buy shows a Ryzen AI 9 365 CPU with a 3K OLED touchscreen for $1,399 USD. That also includes 24GB of memory and a 1TB SSD.

We've seen benchmarks from GPD for its own GPD Duo Dual OLED laptop with the Ryzen AI 9 HX 370, likely a slightly faster option than the 365. The new mobile chips could outpace the desktop Ryzen 9 5950X in CPU performance, with integrated graphics performance slotting between the mobile Nvidia RTX 2050 and RTX 3050 GPUs. Those are only estimates based on 3DMark, however, where the score of 4,221 for the Radeon 890M iGPU reported by GPD indicates a 33% lift in performance over the previous generation Radeon 780M iGPU.

Overall, AMD's upcoming refresh cycle looks pretty enticing, for desktop and laptop users alike. We'll have to see how the laptops stack up against the Qualcomm Snapdragon X Elite solutions when we have hardware in hand, but the rumored performance gains look good while remaining price-competitive with previous solutions.

When AMD announced its Ryzen AI 300-series processors earlier this month, it revealed that they offered around 20% higher CPU performance and approximately 36% higher gaming performance (in undisclosed conditions) compared to predecessors. Apparently, MSI was reportedly a bit more specific about the grunt delivered by AMD's upcoming APUs, and it turns out that their gaming performance is on par with some well-known previous-generation discrete GPUs. However, please keep in mind that we are dealing with unverified information here.

"According to MSI's booth (at Computex? – TH Editor), the Ryzen AI 9 300 series processors have improvements of around 20% in single-core, multi-core, and integrated graphics performance," wrote well-known hardware leaker Golden Pig Upgrade Pack. "This translates to approximately 2,000+ in single-core and 20,000+ in multi-core on the R23 benchmark, and a Time Spy score of around 3,600+."

Assuming that AMD's latest integrated GPU — the Radeon 890M — indeed scores around 3,600 points in 3DMark Time Spy, its performance is more or less in line with the performance of Intel's Arc A370M, GeForce MX 570 for Laptops, or GeForce GTX 1650 for Laptops - or thereabouts.

AMD's flagship Ryzen AI 370 AI HX 370 processor sports 12 general-purpose Zen 5-based CPU cores and an RDNA 3.5-based Radeon 890M GPU with 16 compute units (1,024 stream processors). The new Radeon 890M integrated graphics processor has 33% more stream processors and can indeed deliver significantly higher compute performance compared to the Radeon 780M, its predecessor. 

Yet, as the performance of iGPUs is typically constrained by thermal design power and memory bandwidth, the actual performance increase of iGPUs due to an increased number of SPs is typically lower than the increase of their theoretical compute performance. That said, it is reasonable to expect the Radeon 890M to offer around 20% higher performance than the Radeon 780M in typical use cases. 

Meanwhile, in some games that are not constrained by memory bandwidth but need more compute horsepower, it may well feature significantly higher performance than its predecessor due to the higher number of compute units and architectural innovations of the RDNA 3.5 microarchitecture compared to the RDNA 3 microarchitecture. 

There are a couple of things to note here, though. First, while MSI's performance estimates are a bit more specific than those of AMD, it is likely that the company compared its laptops based on different processors, so the CPUs were not equal (as different laptops use different cooling and memory). Secondly, we have no idea about the drivers used for testing. Thirdly, we cannot verify the leaked information about performance ourselves. In general, while there is good cause to be optimistic, take this leaked information about the performance of AMD's next-generation integrated GPU with a grain of salt.

More performance numbers surrounding the new Ryzen AI processors leaked following AMD's official Computex 2024 announcement. Discovered by Benchleaks, a Ryzen AI 9 HX 370 12-core CPU was benchmarked on Geekbench 6 and sports some very impressive numbers.

The CPU was tested on an Asus ProArt P16 prosumer laptop sporting 64GB of memory. Crucially, the chip was benchmarked using the laptop's silent power plan, which reduces system performance and fan speed to maintain quiet operation even under load. It's worth mentioning that Geekbench 6 classifies the CPU as the "AMD Ryzen AI 9 HX 170" with a 100 series nomenclature. However, this is inaccurate and resulted from AMD switching the branding name to the 300 series at the last minute.

Running in a neutered power state, the Ryzen AI 9 processor scored 2,544 points in the single-core benchmark and 14,158 in the multi-core test. Compared to its predecessor, the Ryzen 9 8940HS, the AI-branded chip is 6.5% faster in the single-core test and 17% faster in the multi-core test. The Ryzen 9 8940HS pumped out scores of 2,380 and 11,775, respectively, for the single and multi-core tests.

The Ryzen AI 9 HX 370's performance is so good that it impressively approaches the performance of AMD's Dragon Range 16-core mobile processors. Looking at the Ryzen 9 7945HX3D benchmark numbers with 3D-VCache technology, the Ryzen AI 9 CPU is only 11% slower in Geekbench's single-core test and 16.2% slower in multi-core performance. The 7945HX3D's Geekbench scores were 2,820 and 16,460 respectively.

These numbers are pretty impressive, all things considered; the fact that we are seeing anything close to Dragon Range levels of performance in an efficiency-optimized power state makes us wonder how much more potent the Ryzen AI 9 HX 370 will be when it is kicked up to its maximum thermal and power envelope. The new chip is part of AMD's latest Ryzen AI 300 series processors sporting its bleeding edge Zen 5 CPU architecture, RDNA 3.5 integrated graphics, and XDNA2 neural engine.

The Ryzen AI 9 HX 370 is a hybrid processor sporting 12 cores in total, consisting of four high-performance Zen 5 cores and eight compact Zen 5c cores. The TDP is rated at 28W, but the configurable TDP ranges between 15W and 54W. As a result, performance will vary wildly depending on what configurable TDP configuration each laptop maker chooses.

We have to take these results with a pinch of salt since Geekbench 6 is just one synthetic test out of many. But, if Geekbench 6's results are indicative of the CPU's real-world performance, AMD's new Zen 5 chips will be really fast, even at lower power envelopes.

AMD unwrapped its new Ryzen AI 300 series, codenamed Strix Point, today at Computex 2024 here in Taipei, Taiwan, with the new chips featuring the new Zen 5 CPU microarchitecture with two types of cores, an upgraded RDNA 3.5 graphics engine, and of course, AMD’s new XDNA 2 engine that enables running AI workloads locally. AMD’s new branding scheme now brings ‘AI’ right into the name of the chip, reflecting the company’s strong focus on its new AI-focused XDNA 2 neural processing unit (NPU) that now offers 50 TOPS of performance, a 5X gain for the third generation of AMD’s AI processors. This level of performance beats all other chips for Windows PCs, including Qualcomm’s promising Snapdragon X Elite. It easily exceeds Microsoft’s 40 TOPS requirement for next-gen AI PCs, allowing Copilot elements to run locally.

AMD has plenty of other advances under the hood, too, with a move to 12-core Zen 5 processors for thin-and-light and ultralight notebooks that used to be limited to eight CPU cores and up to 16 compute units for the new RDNA 3.5 integrated graphics engine, an increase over the previous-gen’s maximum of 12 CUs.

The Ryzen AI 300 series launches with two new models, and as you can see in the slides above, these chips come as a single monolithic die. The flagship Ryzen AI 9 HX 370 has 12 cores and 24 threads that operate at a base of 2.0 GHz and a peak of 5.1 GHz. However, as you can see in the block diagram and listed in a branding slide, the chip has four standard Zen 5 cores and eight density-optimized Zen 5c cores on a single monolithic die with the GPU and NPU cores.

This marks the debut of the smaller Zen 4C cores in the highest-tier Ryzen 9 mobile family, as these cores were previously limited to AMD’s lowest-end Ryzen 5 and 3 models with the prior-gen Hawk Point chips. AMD's Zen 5c cores are designed to consume less space on a processor die than the 'standard' Zen 5 performance cores while delivering enough performance for less demanding tasks, thus saving power and delivering more compute horsepower per square millimeter than was previously possible (deeper dive here). Though this technique is conceptually similar to Intel's E-cores, AMD’s Zen 5c employs the same microarchitecture as the standard Zen 5 cores and supports the same features with its smaller cores. In contrast, Intel’s design employs different architectures and feature support. Still, the smaller Zen 5c cores operate at lower clock rates and thus offer less peak performance than standard cores, but they also preserve die area for other additives, like a larger GPU and NPU.

The standard and density-optimized Zen 5c cores both support threading, but we assume the higher boost clock only applies to the four standard cores (AMD hasn’t shared detailed specs of the Zen 5c cores yet). The HX 370 chips also have 36 MB of total L3 cache, the 50 TOPS XDNA 2 NPU, and the new RDNA 3.5 Radeon 890M graphics engine with 16 CUs that run at 2.9 GHz. The chip has a 28W TDP rating, but its broad cTDP range means that will not reflect its actual operating power level, which we’ll cover below.

The Ryzen AI 9 365 has ten cores, spread among four standard Zen 5 cores and six density-optimized Zen 5C cores, with a base of 2.0 GHz and a peak boost of 5.0 GHz. The chip also has the 50 TOPS NPU and a 12-CU RDNA 3.5 Radeon 880M graphics engine running at 2.9 GHz. Despite the lower CPU and GPU core counts, this chip is also listed with a 28W TDP like its bigger brother, though this rating is now of dubious importance.

AMD’s previous-gen 7040 and 8040 families were comprised of nine models, so the two new Ryzen AI 300 models are obviously just AMD’s first salvo in its new AI-focused lineup. Below, we’ll cover the specs, what we know about the architecture and process node, AMD’s new branding scheme, and AMD’s benchmarks.

Zen 5, Zen 5c, and RDNA 3.5

AMD divulged that its new Zen 5 microarchitecture delivers a 16% increase in IPC over Zen 4. Paired with new process technology (purportedly TSMC 4nm), the new chips should deliver big gains in both performance and power efficiency at any given power level—particularly with the addition of Zen 5c cores. You can read more about the Zen 5 architecture here.

AMD pairs the new CPU cores with its RDNA 3.5 integrated graphics with up to 16 CUs, a notable increase over the previous peak of 12 CUs. AMD hasn’t shared any fine-grained information on the changes with the new graphics architecture, but we expect more information to surface as the chips move closer to market.

Untangling the New Ryzen AI 300 Series Branding

AMD’s new AI branding is an eye-catcher. Aside from using ‘AI’ right in the brand name, AMD also realigned its entire naming scheme, a curious decision given that it had previously announced a rather complex new branding scheme for its mobile lineup that required a decoder ring to navigate. AMD originally said that it expected the new naming scheme to last five years, but given the shift to next-gen AI PCs, the company has now pivoted to AI-specific branding a mere one and a half years later.

In many respects, the new branding is easier to comprehend, at least insofar as it doesn’t require a decoder ring like the previous scheme but is also a bit more ambiguous. AMD’s model numbering reset begins with the 300 series, so named because Strix Point is the third series of AMD’s mobile AI chips with an NPU. The 7040 ‘Phoenix’ and 8040 ‘Hawk Point’ chips were the first two generations, respectively, but the older models will retain their current branding -- the new branding scheme only applies to Strix Point and newer generations.

AMD’s previous-gen chips were carved into swim lanes defined by the TDP, like U-, HS- and H-series (15/28W, 35W, 45W TDPs, respectively), but Strix Point marks a fundamental shift and no longer has a designator for the TDP level. Any Strix Point chip can be assigned from a 15W to a 54W cTDP (configurable TDP) threshold by the OEM, just as with the previous-gen models, but this will now become a more common practice. As such, AMD doesn’t feel the need to call out the TDP rating as a designator in the product name.

I asked AMD how its customers would know which laptops run at a 15W or a 54W TDP despite having the same CPU model name. AMD says this approach is similar to what they are already doing with cTDPs, and the intention is to simplify the buying decision for mainstream buyers at box stores like Best Buy. Unfortunately, this is not entirely unlike Intel’s own policy of not requiring OEMs to divulge cTDP specifications. As such, you’ll need to check reviews to ensure you’re getting a capable machine.

With TDP removed as a branding measuring stick, the new designators merely signify a ‘brand level.’ For instance, the Ryzen AI 9 HX 370 moniker specifies that the chip is a Ryzen 9 model in the HX category, with the HX designator denoting it’s a high-performance top-of-stack model. The last three digits denote the processor number, with the first digit indicating the series (in this case, 300 series) and the last two digits signifying the relative positioning of the chip within that series (for instance, a ‘350’ chip would be lower than the ‘370’). AMD also unveiled new badges with the new Ryzen branding.

AMD XDNA 2 NPU Architecture

AMD shared plenty of details about its new AI-accelerating XDNA 2 neural processing unit (NPU). AMD was the first x86 chipmaker to bring an NPU to both the mobile and desktop PC markets, with its first-gen Phoenix chips delivering 10 TOPS of performance. That then progressed to 16 TOPS with the Hawk Point refresh, mostly due to boosted NPU clock speeds.

AMD’s Strix Point takes AI performance to a whole new level. The re-architected XDNA 2 engine delivers up to 50 TOPS of performance in both INT8, which is how most NPU metrics are spec’d, and Block FP16, a new paradigm we’ll dive into shortly. AMD hasn’t yet shared the full TOPS rating for its chips with the CPU and GPU added in.

Here, we can see the block diagram of the new XDNA 2 engine, and we have the XDNA 1 engine in the following slide. As you can see, there are 3.5x more ‘AI tiles,’ which previously were referred to as AIE-ML units. It appears that the DM units (used for scalar calculations) have been merged into the new AI tile. The new engine is seemingly far more complex and thus performant, but we’ll have to wait for a proper architectural briefing for more details.

AMD’s rearchitecting of its XDNA 2 engine, which was borne of its Xilinx IP when AMD purchased the company, yields a fivefold increase over the first-gen XDNA to 50 TOPs and twice the power efficiency. The enhanced power efficiency is a critical advance, as the primary goal of the NPU is to offload AI tasks to save battery life.

NPUs are typically characterized by performance in INT8 workloads, a less-precise data type that uses less compute and memory to run a model. However, models have to be quantized to the INT8 format and lose a bit of precision in the process. AMD’s XDNA 2 is the first x86 NPU to support Block BF16, a new data format that provides the full accuracy of FP16 with many of the same compute and memory characteristics of INT8. AMD says Block FP16 is plug-and-ply; it doesn’t require quantizing, tuning, or retraining the existing models. 

As such, AMD has the only NPU on the market that supports not only 50 TOPS with INT8 but also a full 50 TOPS with Block FP16. The company showcased a benchmark using both INT8 and Block FP16 to run the same Stable Diffusion prompt to generate an image, the latter of which provided a more accurate image more quickly. It isn’t yet clear if Block FP16 is certified as an IEEE standard.

AMD Ryzen 9 AI 300 Series Benchmarks

As with all vendor-provided benchmarks, AMD’s tests above should be taken with a grain of salt. We’ve also included the test notes at the end of the album.

AMD claims a 5X performance advantage with LLM models over shipping mobile processors in the market today, but notably, the yet-to-be-released 45 TOPS Qualcomm X Elite and 48 TOPS Lunar Lake aren’t included in these benchmarks.

AMD compared to Qualcomm’s publicly posted benchmark data in productivity work, though, claiming anywhere from a 5% advantage in responsiveness to a 10% advantage in productivity workloads with the Ryzen AI 9 HX 370. The company also threw in a whopping 60% performance advantage in graphics performance to underline its gaming chops. Apple’s M3 also makes an appearance, with AMD claiming a 9% advantage in productivity, an 11% advantage in video editing, and a 98% advantage in 3D rendering.

As you would expect, Intel’s Core Ultra 185H also appears in AMD’s benchmark roster, with AMD claiming a 4% lead in productivity workloads, a 40% gain in video editing, and a 73% advantage in 3D rendering. AMD also claims a 36% iGPU performance advantage over the Core Ultra 185H across a spate of titles.

Intel’s upcoming Lunar Lake processors would make an interesting comparison point here, but those haven’t been officially announced yet.

Thoughts

AMD has a full roster of developers with ISVs working with it to enable AI features, with plans for over 150 partners in 2024. The company also has a strong roster of hardware partners, with 100+ platforms starting to arrive next month. That includes many of the usual suspects, like Asus and MSI, both of which have multiple models on display here at Computex 2024.

AMD’s decision to bring its density-optimized Zen 5c cores to its top-tier Ryzen 9 lineup enables it to pack in more compute power in a smaller area of the chip, thus leaving room for a vastly expanded iGPU and NPU, both of which will pay dividends in other facets, like gaming and AI. With the performance benefits of the full-fat Zen 5 cores also under the hood along with a new faster, more efficient process node, the Ryzen AI 300 series chips look to be exceptionally competitive with Intel, Qualcomm, and Apple’s chips, setting the stage for a very competitive 2024 in the mobile market. Now, all that’s left is to see the chips in third-party benchmarks. With the arrival slated for July 2024, we won’t have to wait long to see how these chips stack up.