The battle for gaming supremacy between Intel and AMD has never been as intense as it is now, but AMD has a new ace in the hole. AMD's $449 Ryzen 7 5800X3D uses new cutting-edge 3D-stacked SRAM technology, called 3D V-Cache, to enable a total of 96MB of L3 cache that unlocks tremendous gaming performance, unseating Intel's expensive $738 Core i9-12900KS as the fastest of the Best CPUs for gaming — but at a more forgiving price point. AMD pulled this feat off with an eight-core 16-thread chip based on the same 7nm process and Zen 3 architecture as the original Ryzen 5000 chips that debuted back in 2020, but uses an innovative hybrid bonding technology to fuse an extra slice of cache atop the processing cores, a first for desktop PCs.
The Ryzen 7 5800X3D represents the company's last hurrah for its long-lived Socket AM4 platforms that have shepherded the Ryzen chips from their infancy with the Ryzen 7 1800X in 2017 to their once-dominating position at the top of our CPU benchmark gaming hierarchy last year with the Ryzen 9 5900X.
AMD's chips held the lead in every metric until Intel released its Alder Lake lineup last year, with Intel's Core i9-12900K landing as the fastest gaming CPU we'd ever tested. However, with AMD poised to launch the 5800X3D, Intel attempted to cement itself atop the gaming performance charts with its new Special Edition Core i9-12900KS. That came to market last week with boost speeds reaching up to a blistering 5.5 GHz, a record high for PCs, and for a little over a week, it was the fastest desktop PC chip in all categories.
|Price||Cores | Threads||Base/Boost (GHz)||Total L3 Cache||TDP|
|Ryzen 7 5800X3D||$449||8 | 16||3.4 / 4.5 GHz||96MB||105W|
|Core i9-12900KS||$739||16 Cores / 24 threads||3.4 / 5.5 (P-cores) — 2.5 / 4.0 (E-cores)||30MB||150W / 241W|
|Core i9-12900K / KF||$589 (K) - $564 (KF)||16 Cores / 24 threads||3.2 / 5.2 (P-cores) — 2.4 / 3.9 (E-cores)||30MB||125W / 241W|
Intel's short-lived advantage in gaming came at the cost of extra power, though: The Core i9-12900KS has a 150W processor base power (PBP), a record for a mainstream desktop processor, and we measured up to 300W of power consumption under full load. In contrast, the Ryzen 7 5800X3D has a 105W TDP rating and maxed out at 130W in our tests, showing that it is a far cooler processor that won't require as expensive accommodations, like a beefy cooler, motherboard, and power supply, as the Core i9-12900KS.
The Ryzen 7 5800X3D's 96MB of L3 cache is transparent to the operating system, meaning it doesn't need special accommodations from the OS or software, but it doesn't benefit all games. However, we did see a big uplift in nearly every title we tested.
Here's a quick snapshot of the 5800X3D's average performance in our gaming test suite and key single- and multi-threaded applications. You'll find much more extensive testing below, but this gives you a good general sense of how the Ryzen 7 5800X3D stacks up when it arrives on the market on April 20.
As you can see, the Ryzen 7 5800X3D takes the crown as the fastest gaming chip in our test suite and lives up to AMD's claim that 3D V-Cache delivers an increase in gaming performance equivalent to what we would normally only see from a new microarchitecture. However, the 58000X3D isn't as fast as comparable chips in other types of single- and multi-threaded work beyond gaming. That's because the other models have a core count and frequency advantage. In fact, due to lower clock speeds than its most directly-comparable counterpart, the Ryzen 7 5800X, the 5800X3D is slower in some single-threaded applications.
However, while Intel's 12900KS still delivers leading performance in applications, its hefty $739 premium isn't as good of a buy as AMD's $449 Ryzen 7 5800X3D if you're solely interested in gaming. The same applies to the standard 12900K and 12700K, too, though the Core i7-12700K is a contender if you're looking for a more balanced blend of gaming and application performance around the $410 price point.
Of course, the Ryzen 7 5800X3D is a huge win if you already own a Ryzen system — this chip will drop into almost any AM4 motherboard, saving some cash if you have the right supporting components. Overall, the Ryzen 7 5800X3D is exactly what AMD says it is — a chip optimized specifically for gaming that takes the overall lead.
First, let's take a quick look at the specs, then get right to our full gaming and application test results. Be sure to look for the deep-dive details and testing on the 3D V-Cache, boost frequencies, and thermals later in the article (the latter is particularly interesting).
AMD Ryzen 7 5800X3D Specifications and Pricing
The Ryzen 7 5800X3D is the first consumer processor to feature 3D V-Cache, but the company also uses the tech for its Milan-X processors for the data center. 3D V-Cache leverages a novel new technique that uses hybrid bonding to fuse an additional 64MB of 7nm SRAM cache vertically atop the Ryzen compute chiplet, thus tripling the amount of L3 cache per Ryzen die.
|Street / MSRP||Cores | Threads||P-Core Base/Boost||E-Core Base/Boost||L3 Cache||TDP / PBP / MTP||DDR4-3200|
|Core i9-12900KS||$739||8P + 8E | 16 Cores / 24 threads||3.4 / 5.5 GHz||2.5 / 4.0 GHz||30 MB||150W / 241W||DDR4-3200 / DDR5-4800|
|Core i9-12900K / KF||$589 (K) - $564 (KF)||8P + 8E | 16 Cores / 24 threads||3.2 / 5.2 GHz||2.4 / 3.9 GHz||30MB||125W / 241W||DDR4-3200 / DDR5-4800|
|Ryzen 9 5900X||$450 ($549)||12P | 24 threads||3.7 / 4.8 GHz||-||32MB||105W||DDR4-3200|
|Ryzen 7 5800X3D||$449||8P | 16 threads||3.4 / 4.5 GHz||-||96MB||105W||DDR4-3200|
|Ryzen 7 5800X||$350 ($449)||8P | 16 threads||3.8 / 4.7 GHz||-||32MB||105W||DDR4-3200|
|Core i7-12700K / KF||$409 (K) - $384 (KF)||8P + 4E | 12 Cores / 20 threads||3.6 / 5.0 GHz||2.7 / 3.8 GHz||25MB||125W / 190W||DDR4-3200 / DDR5-4800|
|Ryzen 7 5700X||$299||8P | 16 threads||3.4 / 4.6||-||32MB||65W||DDR4-3200|
The Ryzen 7 5800X3D comes with the same eight Zen 3 cores and 16 threads as the standard Ryzen 7 5800X but has a lower 3.4 GHz base and 4.5 GHz boost frequency within its 105W envelope. AMD trimmed 400 MHz from the base clock and 200 MHz off the boost frequency, but you get an additional 64MB of L3 cache in exchange, for a total of 96MB of L3.
Naturally, the 3D V-Cache tech has tradeoffs, with the most obvious being the $449 price tag — you’ll pay an extra $100 for the same number of cores as you’d get in the vanilla Ryzen 7 5800X.
The 5800X3D's big attraction is AMD's claim of an average 15% gain in gaming performance over AMD's fastest gaming chip, the Ryzen 9 5900X which also currently retails for $450. The 3D V-Cache doesn’t increase performance in other types of work beyond gaming, so compared to the 5900X, you’ll sacrifice four cores and eight threads in exchange for the extra cache, thus losing performance in some productivity applications. That means the Ryzen 9 5900X would be the better option for productivity-focused work, but you should also look to Alder Lake alternatives if you're after a more balanced performance profile.
The 5800X3D fully supports overclocking the memory and Infinity Fabric, but you can't overclock the CPU cores or use the auto-overclocking Precision Boost Overdrive feature (more detail below). The company cites a voltage limitation, but our thermal testing below certainly implies that heat dissipation is an exacerbating issue. AMD says this is the first iteration of the tech, and it is possible that overclocking could be enabled on potential future 3D V-Cache processors. However, the company hasn't officially committed to releasing other models in the future. Given the performance we've seen, it wouldn't be surprising to see this tech carry over into the Zen 4 era.
That won't stop enterprising enthusiasts from trying, though. We've already seen reports of limited BCLK overclocking that can eke out a few hundred extra megahertz (perhaps more on motherboards with external clock generators), and there appears to be a workaround to alter the motherboard's voltage output to the CPU, thus feeding the chip more voltage than AMD intended. Of course, the latter could incur significant risk, but we'll learn more in the coming weeks as enthusiasts put the silicon through the wringer.
As with all other 105W Ryzen 5000 chips, the Ryzen 7 5800X3D doesn't come with a cooler. The chip has the same thickness (Z-height) as all other Ryzen 5000 models, so it is compatible with the broad ecosystem of standard coolers for the AM4 socket. The 5800X3D will drop into existing 400- and 500-series motherboards (Socket AM4), and AMD’s upcoming BIOS updates will also enable support on older 300-series platforms. You'll need a BIOS with AGESA 188.8.131.52b (or newer) for the Ryzen 7 5800X3D.
AMD says that Ryzen 5000 support will vary by vendor, as will the timeline for new BIOS revisions. However, we should see them in the April-May timeframe. Notably, these BIOS revisions will also include the fix for AMD’s fTPM stuttering issues.
The 5800X3D also doesn't support the leading-edge connectivity options, like DDR5 and PCIe 5.0, that you'll find with Alder Lake, but it does support up to DDR4-3200 and PCIe 4.0. AMD won't be able to match intel's connectivity tech until its 5nm Ryzen 7000 ‘Raphael’ Zen 4 CPUs arrive later this year.
3D V-Cache Technology
The idea behind 3D V-Cache is relatively simple, but the execution is complex. The basic idea behind any on-chip cache is to keep frequently accessed data as close to the execution cores as possible, thus eliminating high-latency trips to main memory. As a result, the cores don't have to wait for data, thus staying busier and boosting performance. The L3 cache is slower than other caches (like L1 and L2), but its higher capacity means it can store more data, improving the hit rate (the number of times useful data is held in the cache). There's a reason AMD calls it "Game Cache" — L3 cache is very important to performance, and games, in particular, can suffer from high L3 latency or reduced cache capacity/hit rates.
But a big slab of cache is best. As you can see in the above album, AMD stacks an additional SRAM chiplet, connected via TSVs to the lower die, directly in the center of the compute die (CCD) to isolate it from the heat-generating cores on the sides of the chiplet. However, AMD has to use a silicon shim on top of the cores to create an even surface for the heat spreader that sits atop the chiplet. Contrary to popular belief, this is a single shim that wraps around the chiplet on three sides (images in cache testing section). Silicon is an excellent thermal conductor, but the shim and extra SRAM die will inevitably reduce thermal dissipation from the bottom die, thus resulting in less thermal headroom. We show that impact in our boost frequency and thermal load testing. The extra memory also consumes more power.
AMD says overclocking isn't possible because the cache chiplet and the CCD share the same power plane and the effective voltage limit for the SRAM chiplet weighs in at 1.35V. Since the core voltage can't be altered separately, that prevents overclocking the CPU core frequencies. Unfortunately, this also hampers peak chip frequencies during normal operation, so the 3D V-Cache tech does contribute to the 5800X3D’s lower clock speeds. For perspective, the Ryzen 7 5800X has a 1.5V limit, so it can reach higher clock speeds.
AMD's 3D chip stacking tech is based on TSMC's SoIC technology. TSMC's SoIC is a bumpless chip stacking tech, meaning that it doesn't use microbumps or solder to connect the two die. Instead, the two die are milled to such a perfectly flat surface that the TSV channels can mate without any type of bonding material, reducing the distance between the cache and core by 1000X. That reduces heat and power consumption while boosting bandwidth. You can read much more about the hybrid bonding and manufacturing process here. AMD says the technique uses silicon fab-like manufacturing with back-end like TSVs, which means the production flow is similar to that of a regular chip.
|7nm 3D V-Cache Die||7nm Core Complex Die (CCD)||12nm I/O Die (IOD)|
|Transistor Count||4.7 Billion||4.15 Billion||2.09 Billion|
|MTr/mm^2 (Transistor Density)||~114.6 Million||~51.4 Million||~16.7 Million|
As before, the 7nm Core Complex Die (CCD) has 4.15 billion transistors spread out over 80.7mm^2 of silicon. Meanwhile, the new smaller 7nm 3D V-Cache die measures only 41mm^2, yet has 4.7 billion transistors. As you can see in the table, that means it has slightly more than twice the transistor density, which is due to AMD using a density-optimized version of 7nm that's specialized for SRAM. It's also important to remember that a standard compute die includes several types of transistors (libraries, standard cells) for different purposes, so density varies across the die. In contrast, the V-Cache die uses a largely uniform layout.
The L3 cache chiplet spans the same amount of area as the L3 cache on the CCD underneath, but it also has twice the capacity. That's due to the optimized process, but also partially because the additional L3 cache slice is somewhat 'dumb' — all the control circuitry resides on the base die, which helps reduce the inevitable latency overhead associated with fetching data from a separate die (more on that in the cache testing section later).
Test Setup and Overclocking
As mentioned, the Ryzen 7 5800X3D doesn't support overclocking via the CPU multiplier, so you can't change the core clocks via that method. You also cannot adjust the power limits (PPT, TDC, EDC) or CPU voltage. Additionally, the chip doesn't support the auto-overclocking Precision Boost Overdrive (PBO) feature, and you can't undervolt or underclock.
The 5800X3D fully supports overclocking the memory and Infinity Fabric, but as with most Ryzen chips, we found that we were only able to reach DDR4-3800 with the fabric dialed in at 1900 MHz. This setting allows us to run the memory in the desired low-latency 'coupled' (1:1 ratio) mode. You can get higher with uncoupled memory, but that results in less performance in games.
There have been reports of successful BCLK overclocks with early samples, which ekes out a few hundred extra megahertz of performance. We'll follow up with additional testing as time permits, but be cautious about overclocking benchmarks you might see in the wild: Remember, manipulating the BCLK has been shown in the past to cause inflated benchmark scores with AMD chips — but there are solutions for that.
We test Intel processors with the power limits fully removed for our standard measurements, so the 12900K and 12900KS are running beyond Intel's 'recommended' power settings, but remain within warranty. We haven't yet overclocked the 12900KS fully, so we're subbing in our overclocked 12900K configuration in its place. From what we've seen, it appears that the 12900KS silicon often clocks similarly to its non-S counterparts, but we'll update if we see a meaningful difference.
Aside from a few errant programs for Intel, the overall trends for both AMD and Intel should be similar with Windows 10 and 11. As such, we're sticking with Windows 11 benchmarks in this article. We also stuck with DDR4 for this round of Alder Lake testing, as overall performance trends are generally comparable between DDR4 and DDR5. We have a deeper dive into what that looks like in our initial 12900K review.
We tested the Ryzen 7 5800X3D in two configurations:
- Ryzen 7 5800X3D: Corsair H115i 280mm water cooler, default power limits, DDR4-3200 in Coupled mode
- Ryzen 7 5800X3D DDR4-3800: Corsair H115i 280mm water cooler, default power limits, DDR4-3800 in Coupled mode
AMD Ryzen 7 5800X3D Gaming Benchmarks — The TLDR
As usual, we're testing with an Nvidia GeForce RTX 3090 to reduce GPU-imposed bottlenecks as much as possible, and differences between test subjects will shrink with lesser cards or higher resolutions. Because most of the titles below show little meaningful differentiation at higher resolutions, we only tested four of the seven titles at 1440p. Be aware that the limited selection of titles tested at 1440p can result in large swings in our cumulative measurements if there's a big increase in a single title — those swings would be more muted if we had a larger selection of 1440p titles.
The above charts comprise the geometric mean of our standard gaming test suite, but we include the individual results in the charts below. Given that the 5800X3D's extra cache doesn't benefit all games and that our existing test suite also appears to heavily favor the improvements from 3D V-Cache, we also included a table with results from an additional five games below. Those extra titles aren't factored into the cumulative measurements above, but they show the same general trends.
On average at 1080p, the 5800X3D is ~9% faster than the 12900K, which costs 30% more, and ~7% faster than the Core i9-12900KS, which costs a whopping 64% more. That means the Ryzen 7 58000X3D is now both the fastest gaming chip in our test suite and a better value for gaming specifically than the Core i9 models.
Overclocking either of Intel's Core i9 models requires a beefy cooler and robust motherboard. However, despite its much tamer overall power requirements, the Ryzen 7 5800X3D is still ~3% faster than the overclocked 12900K in our cumulative measurement.
The 5800X3D is 13% faster at 1080p than the stock Core i7-12700K but is only 3.6% faster than the overclocked 12700K config. The Ryzen 7 5800X3D is 10% more expensive than the 12700K, but the more value-centric AM4 ecosystem gives AMD a leg up over Intel's chip, at least if you're specifically interested in gaming. As you'll see in the application testing below, the Core i7-12700K is a much better all-rounder if you're looking for performance in productivity work, too.
AMD's marketing claim is that the Ryzen 7 5800X3D is, on average, 15% faster than the Ryzen 9 5900X. The 3D V-Cache doesn't improve performance in all games, so this will vary, but we recorded a 21% increase over the 5900X at 1080p in our test suite, which is incredibly impressive.
The 5800X3D and the 5800X are built from the same basic design, but the X3D model has a 200 MHz lower boost and 400 MHz lower base clock than the 5800X. Despite that limitation, we recorded a massive 28% gain over the 5800X at 1080p, which is impressive. However, overclocking the 5800X3D's memory yielded an average performance increase of only about 1%, which isn't too meaningful.
|Tom's Hardware - 5800X3D Baseline||1080p Game Benchmarks - fps %age|
|Ryzen 7 5800X3D||100%|
|Core i9-12900KS DDR4||93.5%|
|Core i9-12900K DDR4||91.5%|
|Core i9-12700K DDR4||88.6%|
|Ryzen 9 5900X||82.6%|
|Ryzen 7 5800X||78.1%|
It is noteworthy that a few of our tested game titles approach a GPU bottleneck at 1080p, so we might see larger performance deltas when new, more powerful GPUs arrive later this year. Moving over to 1440p brings a GPU bottleneck into the equation, so the performance deltas between the chips shrink tremendously. However, those results provide a good perspective if you game at higher resolutions and don't plan to upgrade your GPU before buying your next CPU.
The competition between Intel and AMD is much closer now, and not all games benefit from the 3D V-Cache, so it's best to make an informed decision based on the types of titles you play frequently. Be sure to check out the individual tests below.
3DMark, VRMark, Chess Engines on AMD Ryzen 7 5800X3D
Synthetic benchmarks don't tend to translate well to real-world gaming, but they do show us the raw amount of compute power exposed to game engines. It's too bad most games don't fully exploit it. Here we can see that the Ryzen 7 5800X3D is very similar to the 5800X in compute-bound synthetic tests. You'll see much bigger gains in the real-world games below.
Extra AMD Ryzen 7 5800X3D Game Benchmarks - GTA V, Project Cars 3, Shadow of the Tomb Raider, Far Cry 5, Borderlands 3
|Tom's Hardware - 1080p Extras||Ryzen 7 5800X3D||Core i9-12900K||Ryzen 9 5900X||Ryzen 7 5800X|
|Grand Theft Auto V||184.9||187.2||179||176.4|
|Project Cars 3||273.7||257.2||217.5||215.5|
|Shadow of the Tomb Raider||235.7||197.9||174||163.9|
|Far Cry 5||182.8||156.2||123||118.8|
This is, admittedly, not the best way to present this set of test results, but after seeing some of the large deltas in our test suite, we wanted to expand our view to a few more game titles that we don't normally test. Given the time pressure of the NDA lift, we threw together this quick table to give a basic view of a different mix of game titles with stock processor settings. The Ryzen 7 5800X3D comes out ahead in all but Grand Theft Auto V.
Far Cry 6 on AMD Ryzen 7 5800X3D
Far Cry 6 is sensitive to memory latency and throughput. However, it's still surprising that keeping more data close to the processing cores yields a massive 32% speedup over the similarly-equipped Ryzen 7 5800X and a 25% increase over the Ryzen 9 5900X. The gains are more muted against Intel's stable, but the 5800X3D still pulls out the win.
F1 2021 on AMD Ryzen 7 5800X3D
Some game titles respond fantastically to the increased cache capacity, and F1 2021 definitely falls into that category. Here we can see that the stock 5800X3D is 11.6% and 20.4% faster than the stock 12900KS and 12900K, respectively, at 1080p. The 5800X3D is also 20.3% and 25.8% faster than the Ryzen 9 5900X and Ryzen 7 5800X, respectively.
Hitman 3 on AMD Ryzen 7 5800X3D
Intel collaborated with the IO Interactive team to optimize Hitman 3's Glacier 2 game engine for Alder Lake's x86 hybrid architecture, a fact Intel heavily promoted during its launch. Intel takes the lead after overclocking, but the dead-simple stock Ryzen 7 5800X3D setup is 5.5% faster than the stock Core i9-12900KS and 9.6% faster than the 12700K. Those processors are technically faster than the 5800X3D after overclocking, but the 12900K only leads by 1%, while the 12700K pulls off a tie.
Horizon Zero Dawn on AMD Ryzen 7 5800X3D
Horizon Zero Dawn is largely GPU-bottlenecked at the top of the chart, with the overclocked Core i9-12900K and Core i7-12700K marching in lockstep with the Ryzen 7 5800X3D. The Ryzen 7 5800X3D holds a slim lead over the stock Intel configs.
Red Dead Redemption 2 on AMD Ryzen 7 5800X3D
Red Dead Redemption 2 finds the Ryzen 7 5800X3D taking the lead again, though overclocking the memory and Infinity Fabric again gives us no real tangible benefit. The 5800X3D is 9.5% and 11% faster than the stock Core i9-12900K and Core i7-12700K, respectively. Overclocking shrinks the 5800X3D's lead to 5% and 6.5%, respectively.
Watch Dogs Legion on AMD Ryzen 7 5800X3D
Intel's Alder Lake has dominated Watch Dogs Legion since launch, but here we can see that the Ryzen 7 5800X3D carves out yet another impressive win, even in the face of heavily-overclocked Intel challengers.
AMD Ryzen 7 5800X3D Application Benchmarks — The TLDR
We can boil down productivity application performance into two broad categories: single- and multi-threaded. These slides show the geometric mean of performance in several of our most important tests in each category, but be sure to look at the expanded results below.
These results clearly show that the Ryzen 7 5800X3D is a chip designed specifically for gaming, not for leading-edge performance in application workloads. We've highlighted the 5800X3D beating the 12900K in gaming, but we'd be remiss if we didn't mention that the 12900K is 29% faster in single-threaded work and 62% faster in threaded applications. That chasm grows even larger with the Core i9-12900KS.
In fact, even the newly-released Ryzen 7 5700X, which feels like a deep price cut for the 5800X in the disguise of a new product, beats the 5800X3D by 2.5% in threaded work, a byproduct of its 100 MHz higher boost speed. However, the 5700X's 65W TDP is lower than the 5800X3D's 105W, hampering its performance in multi-threaded work, giving the 5800X3D an 11.3% advantage. It'll be interesting to see how the Ryzen 7 5700X stacks up after overclocking in our review next week.
The 5800X3D's lower clock speeds obviously take a toll, as the Ryzen 7 5800X is 7% faster in single-threaded work, but only 1% faster in threaded work.
Given its $409 price point, the Core i7-12700K really shines in comparison to the 5800X3D. The Core i7-12700K is 28.8% faster in single-threaded work and 40% faster in multi-threaded work, showing that it is the best all-rounder in this price range. Yes, that's even after the Ryzen 9 5900X's recent deep price cut to $450.
The Ryzen 7 5800X3D clearly isn't focused on performance in applications outside of gaming, so the below results are fairly predictable. As such, we'll limit our commentary throughout the application benchmarks.
|Tom's Hardware - Application Benchmarks||Single-Threaded||Multi-Threaded|
|Core i9-12900KS DDR4||100%||100%|
|Core i9-12900K DDR4||93.4%||95.9%|
|Core i9-12700K DDR||90.9%||82.6%|
|Ryzen 9 5900X||77.9%||80%|
|Ryzen 7 5800X||77.3%||59.3%|
|Ryzen 7 5800X3D||72%||58.9%|
Rendering Benchmarks on AMD Ryzen 7 5800X3D
The Ryzen 7 5800X3D is capable but underwhelming compared to like-priced competitors throughout this spate of both single- and multi-threaded rendering tests.
|Ryzen 7 5800X3D||106.97||66.67||51.13|
|Ryzen 7 5800X||110.19||67.19||52.43|
|Ryzen 9 5900X||157.79||96.28||74.32|
|Ryzen 7 5700X||95.4||60||45.19|
We've used the Blender command-line utility to automate our performance and power testing for four Blender scenes, building out an extensive library of results over the course of a year...and now the online utility doesn't work and returns an error. We've seen leaked reports of Blender speed-ups from the 3D V-Cache, so we ran a selection of tests with the manual GUI to suss out the differences.
We don't see any tangible benefits over the Ryzen 7 5800X here, and the 5800X3D trails all but the Ryzen 7 5700X.
Encoding Benchmarks on AMD Ryzen 7 5800X3D
It's easy to spot the lightly-threaded encoders in this lineup — the Ryzen 7 5800X3D is at the bottom of the chart for each of them. Without fail, the chip beats the Ryzen 7 5700X in all of the threaded tests. It also exhibits an advantage over the higher-clocked 5800X in the HandBrake tests, implying that the 3D V-Cache confers a slight advantage.
Web Browsing, Office and Productivity on AMD Ryzen 7 5800X3D
The ubiquitous web browser is one of the most frequently used applications. These tests tend to be lightly threaded, so a snappy response time is critical. Again, the 5800X3D lands near the bottom of the test pool in lightly-threaded apps.
Adobe Premiere Pro, Photoshop, and Lightroom on AMD Ryzen 7 5800X3D
We've integrated the UL Benchmarks Procyon tests into our suite to replace the aging PCMark 10. This new benchmark runs complex Adobe Premiere Pro, Photoshop, and Lightroom workflows with the actual software, making for a great real-world test suite.
The Ryzen 7 5800X3D rises to the middle of the pack in the Lightroom and Photoshop benchmark, but it's clear that video editing in Adobe Premiere Pro is not its forte.
Compilation, Compression, AVX Benchmarks on AMD Ryzen 7 5800X3D
This grab bag of various tests finds the Ryzen 7 5800X3D largely trailing its similarly-priced competition, though the increased performance in the threaded y-cruncher test almost certainly results from the 3D V-Cache. However, the 5800X3D's lower clock speeds mean that the same trend doesn't carry over to the lightly-threaded y-cruncher benchmark.
AMD Ryzen 7 5800X3D 3D V-Cache Design and Latency
Several factors influenced AMD's decision to use 3D-stacked SRAM, but key among them is that SRAM density isn't scaling as fast as logic density. As a result, caches now comprise a higher percentage of the die area than before, but without delivering meaningful capacity increases. Furthermore, expanding the cache laterally would incur higher latency due to longer wire lengths and eat into the available die area that AMD could use for cores. Additionally, adding another SRAM chiplet in a 2D layout isn't feasible due to the latency and bandwidth impact.
To address those issues, AMD stacks the additional SRAM directly on top of the center of the compute die where the existing L3 resides. This L3-on-L3 stacking allows the lower die to deliver power and communicate through two rows of TSV connections that extend upwards into the bottom of the L3 cache chiplet. These connections go vertically into the upper die and fan out, which actually reduces the amount of distance data has to travel, thus reducing the number of cycles needed for traversal compared to a standard planar (2D) cache expansion. As a result, the L3 chiplet provides the same 2 TB/s of peak throughput as the on-die L3 cache, but it only comes with a four-cycle latency penalty.
|Tom's Hardware||Ryzen 7 5800X3D||Ryzen 7 5800X|
|AIDA - L3 Latency||13.84 ns||11.49 ns|
|AIDA - L3 Cycles||47 clk||43 clk|
The album above outlines our cache and memory latency benchmarks with the AMD Ryzen 7 5800X3D and the 5800X using the Memory Latency tool from the Chips and Cheese team. These tests measure cache latency with varying sizes of data chunks, and the first slide zooms in on the L3 portion of the cache. Here we can see that the tool measures the Ryzen 7 5800X3D's L3 latency at 12-13ns, whereas the 5800X measures at 10-11ns (the second slide shows the zoomed-out version). We also used AIDA to record the latency measurements, which we listed in the table. Overall, the 3D V-Cache triples the amount of L3 cache but incurs a fairly negligible ~2ns latency impact and a four-cycle penalty.
As mentioned before, the L3 cache chiplet spans the same amount of area as the L3 cache on the CCD underneath, but it has twice the capacity. That's partially because the additional L3 cache slice is somewhat 'dumb' — all the control circuitry resides on the base die, which helps reduce the latency overhead. AMD also uses a density-optimized version of 7nm that's specialized for SRAM. The L3 chiplet is also thinner than the base die (13 metal layers).
AMD produces all of its Zen 3 silicon with TSVs, so all of its Zen 3 silicon supports a 3D V-Cache configuration. However, the TSVs aren't exposed unless they're needed. For 3D V-Cache models, AMD slightly thins the base die as well to both expose the TSV connections and also to maintain the same overall package thickness (Z-Height) as the existing models.
The lack of control circuitry in the L3 chiplet also maximizes capacity and allows AMD to selectively 'light up' only the portions of the cache that are being accessed, thus reducing (and even removing) the power overhead of tripling the L3 cache capacity. In addition, because the larger cache reduces trips to main memory due to higher L3 cache hit rates, the additional capacity relieves bandwidth pressure on main memory, helping to reduce latency and thereby improving application performance from multiple axes. Fewer trips to main memory also reduces overall power consumption.
The L3 cache chiplet consumes significantly less power per square millimeter than the CPU cores. Still, vertical stacking does increase power density, so it's best to isolate it from the heat-generating cores on the sides of the chiplet. However, this would leave a protruding die on top of the CCD, so AMD uses a single silicon shim that wraps around three sides of the L3 chiplet to create an even surface for the heat spreader that sits atop the chiplet. Silicon is an excellent thermal conductor, and the intention is for the shim to allow heat to transfer from the cores up to the heat spreader.
Previous renderings of the design have shown two distinct silicon shims and appeared to show the L3 cache die spanning from one side of the die to the other. However, AMD's materials for the Milan-X launch clearly show one long shim that covers the compute die and a thin portion on the edge of the die that isn't covered by the L3 cache chiplet. This thin expanse of the bottom die includes I/O functions that the chiplet uses to communicate with the I/O die. AMD confirmed that this is the actual layout on all 3D V-Cache processors, like the Ryzen 7 5800X3D, and not the stylized renders shared that show two separate shims.
AMD Ryzen 7 5800X3D Boost Frequencies, Power, and Thermals
Thermal dissipation can limit a chip's peak performance, particularly as process nodes become denser. Adding in the complexity of a 3D-stacked design adds thermal challenges, in this case, exacerbated by the silicon shim stacked atop the CPU cores — this shim transfers heat from the cores to the integrated heat spreader (IHS), but could inevitably reduce the efficiency of the heat transfer from the cores. In effect, it could trap a small amount of heat. Signs of those challenges cropped up in our thermal and boost testing, leading us to conduct more in-depth testing to back up our findings.
The slides above show the Ryzen 7 5800X3D and its nearly-similar counterpart, the Ryzen 7 5800X, which doesn't have a 3D V-Cache design, running through a spate of standard heavily threaded applications (Cinebench, HandBrake, AVX-heavy y-cruncher) to measure power and thermals. We used a Corsair H115i 280mm AIO with the fans cranked to 100% to keep the chips as cool as possible during this test run.
The 5800X3D has a 400 MHz lower base and 200 MHz lower boost clock than the Ryzen 7 5800X, and we can see that the 5800X3D runs at 4.35 GHz during the heaviest multi-core workloads while the 5800X runs at 4.5 GHz. This is expected given the specifications, but we also noticed that the 5800X draws up to 145W while reaching those higher clock speeds, while the 5800X3D only peaks around 120W. This despite both chips having the same 105W TDP and 142W PPT.
Both chips reach the same peak around 80C during the heavy parts of the test, showing that the 5800X3D runs at the same temperature even though the 5800X is consuming 25W more power and running at higher clocks. AMD tells us that thermals aren't the limiting factor that prevents higher clock speeds or the allowance of higher voltages (and thus more heat) for overclocking, but these results certainly imply that the 3D-stacked design doesn't dissipate heat as well as the standard design. To investigate further, we ran a more intense test below.
We turned to a Prime95 test under rigorous conditions to take a closer look at thermal dissipation. We often don't include Prime95 power measurements in our standard CPU reviews, largely because there is a massive disconnect between this extremely rigorous stress test and the power consumption and thermal load generated by most real-world applications. Still, we're specifically looking to push the chips to their throttle point for this test.
To push the processor's limits even harder, we unplugged the fans on the Corsair H115i cooler but left the pump running (unplugging the pump caused clocks to drop too rapidly for our logging to provide any granularity). We then kicked off a Prime95 run with small FFTs, but with AVX instructions disabled. This type of Prime 95 stress test is brutal, and you won't see this type of stress during even the heaviest normal use. As such, remember that we're doing this for science, and not as an indicator of how these chips would function in your PC.
The results clearly show that the 5800X3D peaks at 130W while the 5800X peaks at 145W. Then both chips reach a steady-state temperature of 90C before they begin aggressively throttling power, and thus clock speeds, to remain at this temperature threshold. You'll notice that the 5800X3D encounters this high temperature sooner than the 5800X, and it drops to lower clocks than the 5800X before the end of the test.
Additionally, the 5800X3D remains at 90C while drawing 86W at its lowest point, but the 5800X drew 110W at its lowest point at the same 90C. This shows that the 5800X3D doesn't dissipate as much thermal load as the 5800X within the same temperature threshold. In other words, if all else were equal, the 5800X3D would run hotter than the 5800X under identical conditions. In fact, even when consuming less power, the 5800X3D is undoubtedly hotter than the 5800X.
But to be clear: You would never encounter these conditions during normal use, and the Ryzen 7 5800X3D runs perfectly fine within its specifications. In fact, the 5800X3D is much cooler than the competing Core i9 processors.
Thermal dissipation has been one of the major sticking points that have prevented high-performance 3D chips from going mainstream, but AMD has done an amazing engineering job in bringing thermals under control enough to deliver a chip that provides excellent performance within an acceptable TDP threshold.
Our results certainly heavily imply that thermal dissipation will remain a serious challenge at higher power thresholds, and while AMD contends that voltage is the limiting factor that reigns in the 5800X3D's clock speeds and prohibits overclocking, there could be some room for interpretation of that statement. Most have taken AMD's statement to mean that heat isn't an issue, even though the company has also cited heat as a factor with 3D V-Cache at other times.
As a reminder, voltage, frequency, and thermals are all interrelated. Higher clock frequencies require more voltage, but more voltage results in more heat. A higher voltage may simply push the chip outside its comfortable thermal envelope. As such, 3D V-Cache's 1.35V limit may simply be a product definition designed to keep thermals in check given the chip's thermal design, and not be an actual physical limitation of the technology itself.
As part of our normal test regimen, we tested performance in lightly-threaded work with the H115i cooler. To assess peak boost speeds, we ran through our standard series of lightly-threaded tests (LAME, PCMark10, Geekbench, VRMark, and single-threaded Cinebench).
The 5800X3D reached its peak 4.5 GHz frequency frequently, while the 5800X actually exceeded its 4.7 GHz spec and regularly hit 4.8 GHz. Temperatures and power draw aren't a major concern through most of this test, but there are a series of multi-threaded Geekbench workloads near the 1000-second mark. Again, the 5800X draws more power and runs at higher clocks than the 5800X3D during these periods of heavy load, but it has nearly identical temperatures.
AMD Ryzen 7 5800X3D Power Consumption, Efficiency, and Thermals
AMD's Ryzen chips continue to have excellent power and efficiency metrics. Here we can see that the 5800X3D's position further down the voltage/frequency curve yields excellent results in our Handbrake renders-per-watt efficiency metric.
Here we take a slightly different look at power consumption by calculating the cumulative energy required to perform x264 and x265 HandBrake workloads, respectively. We plot this 'task energy' value in Kilojoules on the left side of the chart.
These workloads are comprised of a fixed amount of work, so we can plot the task energy against the time required to finish the job (bottom axis), thus generating a really useful power chart.
Bear in mind that faster compute times, and lower task energy requirements, are ideal. That means processors that fall the closest to the bottom left corner of the chart are the best. As you can see, the Ryzen 7 5800X3D features a nice blend of power and performance.
3D V-Cache Takes The Lead
The $449 Ryzen 7 5800X3D's 3D V-Cache tech represents an innovative engineering effort that conquered the technical challenges associated with bringing the first desktop PC chip with 3D-stacked SRAM to market, and to great effect. The end result is a comparatively low-power chip that delivers incredible gaming performance, dethroning Intel's $589 Alder Lake Core i9-12900K and $739 Core i9-12900KS from the top of our gaming charts.
The Ryzen 7 5800X3D is a special chip optimized specifically for gaming, but it can't keep pace with similarly-priced chips in productivity applications. The 5800X3D also doesn't support Alder Lake's leading-edge connectivity options, like DDR5 and PCIe 5.0. AMD won't have comparable connectivity until its 5nm Ryzen 7000 ‘Raphael’ Zen 4 CPUs launch later this year.
Below, we have the geometric mean of our gaming test suite at 1080p and 1440p and a cumulative measure of performance in single- and multi-threaded applications. Remember that we conducted the gaming tests with an RTX 3090, so performance deltas will shrink with lesser cards and higher resolution and fidelity settings. However, it is noteworthy that a few of our tested titles are approaching a GPU bottleneck at 1080p, so we might see bigger performance deltas when new, more powerful GPUs arrive later this year.
The 3D V-Cache doesn't accelerate all games, so your mileage will vary. On average in our test suite at 1080p, the 5800X3D is ~9% faster than the 12900K which costs 30% more, and ~7% faster than the Core i9-12900KS which costs a whopping 64% more. The 5800X3D even manages to carve out a 3% lead in the face of the heavily-overclocked 12900KS, making it both the fastest gaming chip in our test suite and a better value for gaming than the Core i9 and Core i7 models.
The 5800X3D also beat AMD's previous-fastest gaming chip, the Ryzen 9 5900X, by 21% in gaming. The 5800X3D does command a sizeable $100 upcharge over the 5800X, but it was 28% faster in our test suite. In fact, despite its lower clock speeds, we didn't see any regressions in gaming with the 5800X3D.
Overall, the Ryzen 7 5800X3D is a great high-performance gaming chip, but its performance characteristics are bipolar when it comes to standard desktop PC applications: The Core i9-12900K is 29% faster in single-threaded work and 62% faster in threaded applications, and the 12900KS is even faster still. Additionally, the Ryzen 7 5800X is 7% faster in single-threaded work, but only 1% faster in multi-threaded applications.
If you're looking for a more balanced chip that does well at both gaming and applications, the $409 Core i7-12700K is a solid choice. The Core i7-12700K is 28.8% faster in single-threaded work and 40% faster in multi-threaded work than the 5800X3D, showing that it is the best all-rounder in this price range. For AMD fans, the $450 Ryzen 9 5900X is also a competent all-arounder, but the 12700K remains a better value.
The 5800X3D drops into existing socket AM4 motherboards dating all the way back to the 300-series that debuted in 2017, so it will make a great high-performance drop-in upgrade for Ryzen owners, provided they have a decent motherboard with solid power delivery and the right supporting components. That upgrade path is even more important given the recent shortages and price hikes we've seen, but it will likely be the last upgrade for socket AM4 platforms. Be aware that AM4 is on the way out to make room for the 5nm Ryzen 7000 ‘Raphael’ Zen 4 CPUs in the AM5 socket. Should you upgrade or wait for Zen 4? Unfortunately, we can't answer that question yet.
AMD hasn't enabled overclocking the 5800X3D's core frequencies, probably for the reasons we outlined in our thermal throttling tests above, but you can tune the memory and Infinity Fabric. Given the 5800X3D's exceptionally high performance in gaming at stock settings, it might be easy for some to forgive this limitation. It's still a bummer, but not a deal-breaker.
The Ryzen 7 5800X3D has much lower power consumption than the Core i9-12900KS and 12900K, making it a far cooler processor that won't require as expensive accommodations, like a beefy cooler, motherboard, and power supply. We are talking about a chip that peaked at 130W compared to the 12900KS that topped out at over 300W, after all. That means the 5800X3D delivers top-notch gaming performance along with a cooler, quieter, and less expensive system than you'll get with a Core i9.
If you're willing to accept the lower but still competent performance in desktop PC applications, and also do your homework to make sure the Ryzen 7 5800X3D accelerates the types of games you play frequently, it's hard to go wrong with this chip — especially for upgraders. Stepping up from a Ryzen 7 1800X to a drop-in Ryzen 7 5800X3D is a no-brainer, for instance. We're sure that will keep plenty of folks from jumping ship to Intel's Alder Lake or upcoming Raptor Lake processors.
Some will balk at the 5800X3D's price tag, and we still recommend that most gamers shop for Core i5 or Ryzen 5 processors for new dedicated gaming rigs. Still, if you have a taste for higher-end fare and a good understanding of the strengths and weaknesses, the Ryzen 7 5800X3D is an impressive chip that delivers leading-edge gaming performance and leaves room for future GPU upgrades, joining our list of the Best CPUs for gaming as the best high-end value for gaming.
|Intel Socket 1700 DDR4 (Z690)||Core i9-12900KS, Core i9-12900K, Core i7-12700K|
|MSI Z690A WiFi DDR4|
|2x 8GB Trident Z Royal DDR4-3600 - Stock: DDR4-3200 14-14-14-36 / OC: DDR4-3800 - All Gear 1|
|AMD Socket AM4 (X570)||AMD Ryzen 7 5800X3D, Ryzen 9 5900X, Ryzen 7 5800X, Ryzen 7 5700X|
|ASUS ROG Crosshair VIII Dark Hero|
|2x 8GB Trident Z Royal DDR4-3600 - Stock: DDR4-3200 14-14-14-36 | OC/PBO: DDR4-3800|
|All Systems||Gigabyte GeForce RTX 3090 Eagle - Gaming and ProViz applications|
|Nvidia GeForce RTX 2080 Ti FE - Application tests|
|2TB Sabrent Rocket 4 Plus - Silverstone ST1100-TI - Corsair H115i AIO - Arctic MX-4 TIM - Open Benchtable - Windows 11 Pro|