AMD Ryzen 9 7900X test: A BANG of an intergenerational leap

Graphics effects: Adobe After Effects

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

AMD Ryzen 9 7900X in detail

AMD Ryzen 7000 processors bring a new manufacturing node, a new architecture and a new platform over the previous generation. The Raphael codenamed models, of which we’ll now be testing the Ryzen 9 7900X, are made from chiplets. The I/O chiplet containing the connectivity and memory controller is manufactured by TSMC’s 6nm process (N5).

Processors with eight or six cores use a single chiplet (CCD), and processors with 12 and 16 cores have two. This basic build is unchanged from the Ryzen 5000, but the cooling is significantly more demanding. That’s due to the reduction in chiplet area with CPU cores, from 80.7 mm² to 71 mm². While this is not a significant reduction (it is still “only” 12 %), there is a significant increase in TDP on top of it. For the same power draw, the increase in temperatures is quite small, but when you add in the significantly higher power draw, cooling Ryzen 7000s below 95°C is a challenge even with powerful coolers (if they are to be quiet). The higher end Ryzen 9 models have a TDP of 170 W and the PPT figure has risen to 230 W (so 11 W below the PL2 of the “K” Core i9 Alder Lake).

The main update is the new Zen 4 architecture. This is developed as an evolutionary improvement on the Zen 3 core, which was the new architecture. AMD uses a sort of two-step cycle in this, where one architecture is always new and the next one then builds on it, only to have a more significant rebuild come along again. The next significantly new architecture will be Zen 5, from which in turn the evolutionarily improved Zen 6 will be derived. According to AMD, Zen 4 has on average 13 % better IPC (performance at identical clock frequency) than Zen 3. However, the increase in IPC may be different in each application, as the percentage is based on the geometric mean of the selected tasks.

AMD hasn’t specified what the exact changes to the architecture are yet. We covered what is known in this article. For example, the uOP cache in the core has been increased, and perhaps the most noticeable change is the 1-megabyte L2 cache (compared to Zen 1 to 3, this is double their capacity).

A big factor in the improvement of the final performance are especially the clock speeds, which shoot upwards really dramatically. The Ryzen 9’s all-core boost in games even at lower (typically gaming) workloads exceeds 5.4 GHz, which puts it 800MHz above its predecessor (Ryzen 9 5900X) as well as the single-core boost at 5.6 GHz. With that one, AMD also breaks through the ceiling very significantly. For comparison, the Core i9-11900K, as the processor with the second fastest ST boost, achieves 300 MHz less. The speed increase of the Ryzen 9 7900X (and the Ryzen 7000 overall) is thus noticeable even in activities that use the performance of only one core at a time.

The Zen 4 core architecture brings one more rather significant change – for the first time, AVX-512 instructions with 512-bit vectors are supported in AMD processors. Their eventual contribution will be discussed in a separate test, which will be separated from the set of standard measurements. However, it is already worth noting that the implementation uses 256-bit units, which Zen 3 already had, and the 512-bit vector is processed on two passes. According to AMD, AVX-512 computed with 256-bit units has one advantage – there will be no underclocking of the cores when these instructions are computed, which has plagued Intel’s implementation (Rocket Lake). In addition to AVX-512, Zen 4 is the first AMD processor to support VNNI instructions.

Pinless processors, DDR5 and PCIe 5.0

There are more of those things that premiere with Ryzen 7000 processors. One of them is AMD’s new AM5 platform, which uses a motherboard contact plate socket (LGA 1718). The pins on AMD processors end with Zen 4. With these processors and motherboards, DDR5 memory support (officially DDR5-5200, unofficially more) is coming to the AMD platform at the same time, DDR4 is not supported. AMD processors do not have (like Intel, for example) a dual memory controller that would also support the older standard (DDR4). Ryzen 7000s only support DDR5 memory.

Also new is PCI Express 5.0 support. The Ryzen 7000 provides this for the graphics card (PCI Express 5.0 ×16) and SSD – the processors theoretically have a pair of PCIe 5.0 ×16 interfaces to directly connect the NVMe SSD to the processor, but both do not have to be brought out to slots on the motherboard.

The presence of an integrated graphics core can also be considered a rather significant change compared to previous generations. AMD’s more powerful processors for the AM4 socket did not have iGPUs (and it only concerned APUs), but the Ryzen 7000 “Raphael” now have a small GPU integrated in the I/O chiplet with two CUs (128 shaders) of the RDNA 2 architecture. The gaming performance of this graphics isn’t whopping, but it will allow you to hook up a monitor and function without a graphics card similar to Intel processors. With this, AMD finally erases one of the competition’s advantages.

All the parameters of the Ryzen 9 (7900X), measurements of which occupy the following chapters of the article, are available in the table below. It compares with the Intel Core i9-12900K, which is closest in price to AMD’s 12-core (and 24-thread) processor.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Gaming tests

We test performance in games in four resolutions with different graphics settings. To warm up, there is more or less a theoretical resolution of 1280 × 720 px. We had been tweaking graphics settings for this resolution for a long time. We finally decided to go for the lowest possible (Low, Lowest, Ultra Low, …) settings that a game allows.
One could argue that a processor does not calculate how many objects are drawn in such settings (so-called draw calls). However, with high detail at this very low resolution, there was not much difference in performance compared to FHD (which we also test). On the contrary, the GPU load was clearly higher, and this impractical setting should demonstrate the performance of a processor with the lowest possible participation of a graphics card.
At higher resolutions, high settings (for FHD and QHD) and highest (for UHD) are used. In Full HD it’s usually with Anti-Aliasing turned off, but overall, these are relatively practical settings that are commonly used.

The selection of games was made considering the diversity of genres, player popularity and processor performance requirements. For a complete list, see Chapters 7–16. A built-in benchmark is used when a game has one, otherwise we have created our own scenes, which we always repeat with each processor in the same way. We use OCAT to record fps, or the times of individual frames, from which fps are then calculated, and FLAT to analyze CSV. Both were developed by the author of articles (and videos) from GPUreport.cz. For the highest possible accuracy, all runs are repeated three times and the average values of average and minimum fps are drawn in the graphs. These multiple repetitions also apply to non-gaming tests.

Computing tests

Let’s start lightly with PCMark 10, which tests more than sixty sub-tasks in various applications as part of a complete set of “benchmarks for a modern office”. It then sorts them into fewer thematic categories and for the best possible overview we include the gained points from them in the graphs. Lighter test tasks are also represented by tests in a web browser – Speedometer and Octane. Other tests usually represent higher load or are aimed at advanced users.
We test the 3D rendering performance in Cinebench. In R20, where the results are more widespread, but mainly in R23. Rendering in this version takes longer with each processor, cycles of at least ten minutes. We also test 3D rendering in Blender, with the Cycles render in the BMW and Classroom projects. You can also compare the latter with the test results of graphics cards (contains the same number of tiles).
We test how processors perform in video editing in Adobe Premiere Pro and DaVinci Resolve Studio 17. We use a PugetBench plugin, which deals with all the tasks you may encounter when editing videos. We also use PugetBench services in Adobe After Effects, where the performance of creating graphic effects is tested. Some subtasks use GPU acceleration, but we never turn it off, as no one will do it in practice. Some things don’t even work without GPU acceleration, but on the contrary, it’s interesting to see that the performance in the tasks accelerated by the graphics card also varies as some operations are still serviced by the CPU.

We test video encoding under SVT-AV1, in HandBrake and benchmarks (x264 HD and HWBot x265). x264 HD benchmark works in 32-bit mode (we did not manage to run 64-bit consistently on W10 and in general on newer OS’s it may be unstable and show errors in video). In HandBrake we use the x264 processor encoder for AVC and x265 for HEVC. Detailed settings of individual profiles can be found in the corresponding chapter 25. In addition to video, we also encode audio, where all the details are also stated in the chapter of these tests. Gamers who record their gameplay on video can also have to do with the performance of processor encoders. Therefore, we also test the performance of “processor broadcasting” in two popular applications OBS Studio and Xsplit.
We also have two chapters dedicated to photo editing performance. Adobe has a separate one, where we test Photoshop via PugetBench. However, we do not use PugetBench in Lightroom, because it requires various OS modifications for stable operation, and overall we rather avoided it (due to the higher risk of complications) and create our own test scenes. Both are CPU intensive, whether it’s exporting RAW files to 16-bit TIFF with ProPhotoRGB color space or generating 1:1 thumbnails of 42 lossless CR2 photos.
However, we also have several alternative photo editing applications in which we test CPU performance. These include Affinity Photo, in which we use a built-in benchmark, or XnViewMP for batch photo editing or ZPS X. Of the truly modern ones, there are three Topaz Labz applications that use AI algorithms. DeNoise AI, Gigapixel AI and Sharpen AI. Topaz Labs often and happily compares its results with Adobe applications (Photoshop and Lightroom) and boasts of better results. So we’ll see, maybe we’ll get into it from the image point of view sometime. In processor tests, however, we are primarily focused on performance.

We test compression and decompression performance in WinRAR, 7-Zip and Aida64 (Zlib) benchmarks, decryption in TrueCrypt and Aida64, where in addition to AES there are also SHA3 tests. In Aida64, we also test FPU in the chapter of mathematical calculations. From this category you may also be interested in the results of Stockfish 13 and the number of chess combinations achieved per unit time. We perform many tests that can be included in the category of mathematics in SPECworkstation 3.1. It is a set of professional applications extending to various simulations, such as LAMMPS or NAMD, which are molecular simulators. A detailed description of the tests from SPECworkstation 3.1 can be found at spec.org. We do not test 7-zip, Blender and HandBrake from the list for redundancy, because we test performance in them separately in applications. A detailed listing of SPECWS results usually represents times or fps, but we graph “SPEC ratio”, which represents gained points—higher means better.

Processor settings…

We test processors in the default settings, without active PBO2 (AMD) or ABT (Intel) technologies, but naturally with active XMP 2.0.

… and app updates

The tests should also take into account that, over time, individual updates may affect performance comparisons. Some applications are used in portable versions, which are not updated or can be kept on a stable version, but this is not the case for some others. Typically, games update over time. On the other hand, even intentional obsolescence (and testing something out of date that already behaves differently) would not be entirely the way to go.
In short, just take into account that the accuracy of the results you are comparing decreases a bit over time. To make this analysis easier for you, we indicate when each processor was tested. You can find this in the dialog box, where there is information about the test date of each processor. This dialog box appears in interactive graphs, just hover the mouse cursor over any bar.



AMD’s first processor with 3D V-cache is a rather controversial piece of hardware. Sure, it may be the ultimate in gaming performance, but the desired effect is more “on paper” than practical, and when it does come, it’s in very rare cases. So that you don’t end up disappointed with a virtually single-purpose processor that may not even excel at gaming, we’ve broken it all down in detailed tests.

Methodology: how we measure power draw

Measuring CPU power consumption is relatively simple, much easier than with graphics cards. All power goes through one or two EPS cables. We also use two to increase the cross-section, which is suitable for high performance AMD processors up to sTR(X)4 or for Intel HEDT, and in fact almost for mainstream processors as well. We have Prova 15 current probes to measure current directly on the wires. This is a much more accurate and reliable way of measuring than relying on internal sensors.
The only limitation of our current probes may be when testing the most powerful processors. These already exceed the maximum range of 30 A, at which high accuracy is guaranteed. For most processors, the range is optimal (even for measuring a lower load, when the probes can be switched to a lower and more accurate range of 4 A), but we will test models with power consumption over 360 W on our own device, a prototype of which we have already built. Its measuring range will no longer be limiting, but for the time being we will be using the Prova probes in the near future.

The probes are properly set to zero and connected to a UNI-T UT71E multimeter before each measurement. It records samples of current values during the tests via the IR-USB interface and writes them in a table at one-second intervals. We can then create bar graphs with power consumption patterns. But we always write average values in bar graphs. Measurements take place in various load modes. The lowest represents an idle Windows 10 desktop. This measurement takes place on a system that had been idle for quite some time.

Audio encoding (FLAC) represents a higher load, but processors use only one core or one thread for this. Higher loads, where more cores are involved, are games. We test power consumption in F1 2020, Shadow of the Tomb Raider and Total War Saga: Troy in 1920 × 1080 px. In this resolution, the power consumption is usually the highest or at least similar to that in lower or higher resolutions, where in most cases the CPU power draw rather decreases due to its lower utilization.
Like most motherboard manufacturers, we too ignore the time limit for “Tau”, after which the power consumption is to be reduced from the PL2 boost limit (when it exceeds the TDP) to the TDP/PL1 value, recommended by Intel, in our tests. This means that neither the power draw nor the clock speed after 56 seconds of higher load does not decrease and the performance is kept stable with just small fluctuations. We had been considering whether or not to respect the Tau. In the end, we decided not to because the vast majority of users won’t either, and therefore the results and comparisons would be relatively uninteresting. The solution would be to test with and without a power limit, but this is no longer possible due to time requirements. We will pay more attention to the behavior of PL2 in motherboard tests, where it makes more sense.
We always use motherboards with extremely robust, efficient VRM, so that the losses on MOSFETs distort the measured results as little as possible and the test setups are powered by a high-end 1200 W BeQuiet! Dark Power Pro 12 power supply. It is strong enough to supply every processor, even with a fully loaded GeForce RTX 3080, and at the same time achieves above-standard efficiency even at lower load. For a complete overview of test setup components, see Chapter 5 of this article.

AMD’s first processor with 3D V-cache is a rather controversial piece of hardware. Sure, it may be the ultimate in gaming performance, but the desired effect is more “on paper” than practical, and when it does come, it’s in very rare cases. So that you don’t end up disappointed with a virtually single-purpose processor that may not even excel at gaming, we’ve broken it all down in detailed tests.

Methodology: temperature and clock speed tests

When choosing a cooler, we eventually opted for Noctua NH-U14S. It has a high performance and at the same time there is also the TR4-SP3 variant designed for Threadripper processors. It differs only by the base, the radiator is otherwise the same, so it will be possible to test and compare all processors under the same conditions. The fan on the NH-U14S cooler is set to a maximum speed of 1,535 rpm during all tests.
Measurements always take place on a bench-wall in a wind tunnel which simulates a computer case, with the difference that we have more control over it.
System cooling consists of four Noctua NF-S12A PWM fans, which are in an equilibrium ratio of two at the inlet and two at the outlet. Their speed is set at a fixed 535 rpm, which is a relatively practical speed that is not needed to be exceeded. In short, this should be the optimal configuration based on our tests of various system cooling settings.

It is also important to maintain the same air temperature around the processors. Of course, this also changes with regard to how much heat a particular processor produces, but at the inlet of the tunnel it must always be the same for accurate comparisons. In our air-conditioned test lab, it is currently in the range of 21–21.3 °C.
Maintaining a constant inlet temperature is necessary not only for a proper comparison of processor temperatures, but especially for unbiased performance comparisons. Trend of clock speed and especially single-core boost depends on the temperature. In the summer at higher temperatures, processors may be slower in living spaces than in the winter.

For Intel processors, we register the maximum core temperature for each test, usually of all cores. These maximum values are then averaged and the result is represented by the final value in the graph. From the outputs of single-threaded load, we only pick the registered values from active cores (these are usually two and alternate during the test). It’s a little different with AMD processors. They don’t have temperature sensors for every core. In order for the procedure to be as methodically as possible similar to that applied on Intel processors, the average temperature of all cores is defined by the highest value reported by the CPU Tdie sensor (average). For single-threaded load, however, we already use a CPU sensor (Tctl/Tdie), which usually reports a slightly higher value, which better corresponds to the hotspots of one or two cores. But these values as well as the values from all internal sensors must be taken with a grain of salt, the accuracy of the sensors varies across processors.
Clock speed evaluation is more accurate, each core has its own sensor even on AMD processors. Unlike temperatures, we plot average clock speed values during tests in graphs. We monitor the temperature and clock speed of the processor cores in the same tests, in which we also measure the power consumption. And thus, gradually from the lowest load level on the desktop of idle Windows 10, through audio encoding (single-threaded load), gaming load in three games (F1 2020, Shadow of the Tomb Raider and Total War Saga: Troy), to a 10-minute load in Cinebench R23 and the most demanding video encoding with the x264 encoder in HandBrake.
To record the temperatures and clock speed of the processor cores, we use HWiNFO, in which sampling is set to two seconds. With the exception of audio encoding, the graphs always show the averages of all processor cores in terms of temperatures and clock speed. During audio encoding, the values from the loaded core are given.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Test setup

* We use the following BIOSes on motherboards. For MSI MEG X670E Ace v1.10NPRP, for MEG X570 Ace v1E, for MEG Z690 Unify v10, for MAG Z690 Tomahawk WiFi DDR4 v11, for MEG Z590 Ace v1.14 and for MEG Z490 Ace v17.

Note: The graphics drivers we use are Nvidia GeForce 466.77 and the Windows 10 OS build is 19043 at the time of testing.

Intel processors are tested on MSI MEG Z690 Unify, MAG Z490 Tomahawk WiFi DDR4, Z590 Ace and Z490 Ace motherboards. Kingston Fury Beast DDR5 memory (2×16 GB, 5200 MHz/CL40) is used with the MSI MEG Z690 Unify.

On platforms supporting DDR5 memory, we use two different sets of modules. For more powerful processors with “X” (AMD) or “K” (Intel) in the name, the faster G.Skill Trident Z5 Neo (2×16 GB, 6000 MHz/CL30) memory. In the case of cheaper processors (without X or K at the end of the name), the slower Kingston Fury Beast (2×16 GB, 5200 MHz/CL40) modules. But this is more or less just symbolism, the bandwidth is very high for both kits, it is not a bottleneck, and the difference in processor performance is very small, practically negligible, across the differently fast memory kits.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

3DMark

We use 3DMark Professional for the tests and the following tests: Night Raid (DirectX 12), Fire Strike (DirectX 11) and Time Spy (DirectX 12). In the graphs you will find partial CPU scores, combined scores, but also graphics scores. You can find out to what extent the given processor limits the graphics card.

---

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Assassin’s Creed: Valhalla

Test environment: resolution 1280 × 720 px; graphics settings preset Low; API DirectX 12; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 1920 × 1080 px; graphics settings preset Low; API DirectX 12; extra settings Anti-Aliasing: low; test scene: built-in benchmark.

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 12; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Ultra High; API DirectX 12; no extra settings; test scene: built-in benchmark.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Borderlands 3

Test environment: resolution 1280 × 720 px; graphics settings preset Very Low; API DirectX 12; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 1920 × 1080 px; graphics settings preset High; API DirectX 12; extra settings Anti-Aliasing: None; test scene: built-in benchmark.

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 12; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Ultra; API DirectX 12; no extra settings; test scene: built-in benchmark.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Counter-Strike: GO

Test environment: resolution 1280 × 720 px; lowest graphics settings and w/o Anti-Aliasing, API DirectX 9; test platform script with Dust 2 map tour.

---

Test environment: resolution 1920 × 1080 px; high graphics settings and w/o Anti-Aliasing, API DirectX 9; test platform script with Dust 2 map tour.

---

Test environment: resolution 2610 × 1440 px; high graphics settings; 4× MSAA, API DirectX 9; test platform script with Dust 2 map tour.

---

Test environment: resolution 3840 × 2160 px; very high graphics settings; 4× MSAA, API DirectX 9; test platform script with Dust 2 map tour.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Cyberpunk 2077

Test environment: resolution 1280 × 720 px; graphics settings preset Low; API DirectX 12; no extra settings; test scene: custom (Little China).

---

Test environment: resolution 1920 × 1080 px; graphics settings preset High; API DirectX 12; no extra settings; test scene: custom (Little China).

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 12; no extra settings; test scene: custom (Little China).

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Ultra; API DirectX 12; no extra settings; test scene: custom (Little China).

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

DOOM Eternal

Test environment: resolution 1280 × 720 px; graphics settings preset Low; API Vulkan; extra settings Present From Compute: off, Motion Blur: Low, Depth of Field Anti-Aliasing: off; test scene: custom.

---

Test environment: resolution 1920 × 1080 px; graphics settings preset High; API Vulkan; extra settings Present From Compute: on, Motion Blur: High, Depth of Field Anti-Aliasing: off; test scene: custom.

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API Vulkan; extra settings Present From Compute: on, Motion Blur: High, Depth of Field Anti-Aliasing: on; test scene: custom.

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Ultra Nightmare; API Vulkan; extra settings Present From Compute: on, Motion Blur: High, Depth of Field Anti-Aliasing: on; test scene: custom.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

F1 2020

Test environment: resolution 1280 × 720 px; graphics settings preset Ultra Low; API DirectX 12; extra settings Anti-Aliasing: off, Anisotropic Filtering: off; test scene: built-in benchmark (Australia, Clear/Dry, Cycle).

---

Test environment: resolution 1920 × 1080 px; graphics settings preset High; API DirectX 12; extra settings Anti-Aliasing: off, Skidmarks Blending: off; test scene: built-in benchmark (Australia, Clear/Dry, Cycle).

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 12; extra settings Anti-Aliasing: TAA, Skidmarks Blending: off; test scene: built-in benchmark (Australia, Clear/Dry, Cycle).

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Ultra High; API DirectX 12; extra settings Anti-Aliasing: TAA, Skidmarks Blending: off; test scene: built-in benchmark (Australia, Clear/Dry, Cycle).

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Metro Exodus

Test environment: resolution 1280 × 720 px; graphics settings preset Low; API DirectX 12; no extra settings test scene: built-in benchmark.

---

Test environment: resolution 1920 × 1080 px; graphics settings preset High; API DirectX 12; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 12; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Extreme; API DirectX 12; no extra settings; test scene: built-in benchmark.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Microsoft Flight Simulator

Disclaimer: The performance of this game changes and improves frequently due to continuous updates. We verify the consistency of the results by re-testing the Ryzen 9 5900X processor before each measurement. In case of significant deviations, we discard the older results and start building the database from scratch. Due to the incompleteness of the MFS results, we do not use MFS to calculate the average gaming performance of the processors.

Test environment: resolution 1280 × 720 px; graphics settings preset Low; API DirectX 11; extra settings Anti-Aliasing: off; test scene: custom (Paris-Charles de Gaulle, Air Traffic: AI, February 14, 9:00) autopilot: from 1000 m until hitting the terrain.

---

Test environment: resolution 1920 × 1080 px; graphics settings preset Low; API DirectX 11; extra settings Anti-Aliasing: off; test scene: custom (Paris-Charles de Gaulle, Air Traffic: AI, February 14, 9:00) autopilot: from 1000 m until hitting the terrain.

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 11; extra settings Anti-Aliasing: TAA; test scene: custom (Paris-Charles de Gaulle, Air Traffic: AI, February 14, 9:00) autopilot: from 1000 m until hitting the terrain.

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Ultra; API DirectX 11; extra settings Anti-Aliasing: TAA; test scene: custom (Paris-Charles de Gaulle, Air Traffic: AI, February 14, 9:00) autopilot: from 1000 m until hitting the terrain.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Shadow of the Tomb Raider

Test environment: resolution 1280 × 720 px; graphics settings preset Lowest; API DirectX 12; extra settings Anti-Aliasing: off; test scene: built-in benchmark.

---

Test environment: resolution 1920 × 1080 px; graphics settings preset High; API DirectX 12; extra settings Anti-Aliasing: off; test scene: built-in benchmark.

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 12; extra settings Anti-Aliasing: TAA; test scene: built-in benchmark.

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Highest; API DirectX 12; extra settings Anti-Aliasing: TAA; test scene: built-in benchmark.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Total War Saga: Troy

Test environment: resolution 1280 × 720 px; graphics settings preset Low; API DirectX 11; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 1920 × 1080 px; graphics settings preset High; API DirectX 11; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 2610 × 1440 px; graphics settings preset High; API DirectX 11; no extra settings; test scene: built-in benchmark.

---

Test environment: resolution 3840 × 2160 px; graphics settings preset Ultra; API DirectX 11; no extra settings; test scene: built-in benchmark.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Overall gaming performance

To calculate average gaming performance, we normalized the Intel Core i7-11900K processor. The percentage differences of all other processors are based on this, with each of the games contributing an equal weight to the final result. To see exactly what the formula we use to arrive at each value looks like, see „New average CPU score measuring method“.

---

---

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Gaming performance per euro

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

PCMark

Geekbench

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Speedometer (2.0) and Octane (2.0)

Test environment: We’re using a portable version of Google Chrome (91.0.472.101) 64-bit so that real-time results are not affected by browser updates. GPU hardware acceleration is enabled as each user has in the default settings.

---

Note: The values in the graphs represent the average of the points obtained in the subtasks, which are grouped according to their nature into seven categories (Core language features, Memory and GC, Strings and arrays, Virtual machine and GC, Loading and Parsing, Bit and Math operations and Compiler and GC latency).

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Cinebench R20

Cinebench R23

---

Blender@Cycles

Test environment: We use well-known projects BMW (510 tiles) and Classroom (2040 tiles) and renderer Cycles. Render settings are set to None, with which all the work falls on the CPU.

---

LuxRender (SPECworkstation 3.1)

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Adobe Premiere Pro (PugetBench)

Test environment: set of PugetBench tests. App version of Adobe Premiere Pro is 15.2.

---

---

---

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

DaVinci Resolve Studio (PugetBench)

Test environment: set of PugetBench tests, test type: standard. App version of DaVinci Resolve Studio is 17.2.1 (build 12).

---

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Graphics effects: Adobe After Effects

Test environment: set of PugetBench tests. App version of Adobe After Effects is 18.2.1.

---

---

---

---

---

---

---

---

---

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

HandBrake

Test environment: For video conversion we’re using a 4K video LG Demo Snowboard with a 43,9 Mb/s bitrate. AVC (x264) and HEVC (x265) profiles are set for high quality and encoder profiles are “slow”. HandBrake version is 1.3.3 (2020061300).

x264 and x265 benchmarks

---

SVT-AV1

Test environment: We are encoding a short, publicly available sample park_joy_2160p50.y4m: uncompressed video 4096 × 2160 px, 8bit, 50 fps. Length is 500 frames with encoding quality set to 6 which makes the encoding still relatively slow. This test can make use of the AVX2 i AVX-512 instructions.

Version: SVT-AV1 Encoder Lib v0.8.7-61-g685afb2d via FFMpeg N-104429-g069f7831a2-20211026 (64bit)
Build from: https://github.com/BtbN/FFmpeg-Builds/releases
Command line: ffmpeg.exe -i “park_joy_2160p50.y4m” -c:v libsvtav1 -rc 0 -qp 55 -preset 6 -f null output.webm

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Audio encoding

Test environment: Audio encoding is done using command line encoders, we measure the time it takes for the conversion to finish. The same 42-minute long 16-bit WAV file (stereo) with 44.1 kHz is always used (Love Over Gold by Dire Straits album rip in a single audio file).

Encoder settings are selected to achieve maximum or near maximum compression. The bitrate is relatively high, with the exception of lossless FLAC of about 200 kb/s.

Note: These tests measure single-thread performance.

FLAC: reference encoder 1.3.2, 64-bit build. Launch options: flac.exe -s -8 -m -e -p -f

MP3: encoder lame3.100.1, 64-bit build (Intel 19 Compiler) from RareWares. Launch options: lame.exe -S -V 0 -q 0

AAC: uses Apple QuickTime libraries, invoked through the application from the command line, QAAC 2.72, 64-bit build, Intel 19 Compiler (does not require installation of the whole Apple package). Launch options: qaac64.exe -V 100 -s -q 2

Opus: reference encoder 1.3.1, Launch options: opusenc.exe –comp 10 –quiet –vbr –bitrate 192

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Broadcasting

Test environment: Applications OBS Studio and Xsplit. We’re using the built-in benchmark (scene Australia, Clear/Dry, Cycle) in F1 2020, in a resolution of 2560 × 1440 px and the same graphics settings, as with standard game performance tests. Thanks to this, we can measure the performance decrease if you record your gameplay with the x264 software encoder while playing. The output is 2560 × 1440 px at 60 fps.

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Adobe Photoshop (PugetBench)

Test environment: set of PugetBench tests. App version of Adobe Photoshop is 22.4.2.

---

---

---

Adobe Lightroom Classic

Test environment: With the settings above, we export 42 uncompressed .CR2 (RAW Canon) photos with a size of 20 Mpx. Then we create 1:1 previews from them, which also represent one of the most processor intensive tasks in Lightroom. The version of Adobe Lightroom Classic is 10.3

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Affinity Photo (benchmark)

Test environment: built-in benchmark.

---

Topaz Labs AI apps

Topaz DeNoise AI, Gigapixel AI and Sharpen AI. These single-purpose applications are used for restoration of low-quality photos. Whether it is high noise (caused by higher ISO), raster level (typically after cropping) or when something needs extra focus. The AI performance is always used.

Test environment: As part of batch editing, 42 photos with a lower resolution of 1920 × 1280 px are processed, with the settings from the images above. DeNoise AI is in version 3.1.2, Gigapixel in 5.5.2 and Sharpen AI in 3.1.2.

XnViewMP

Test environment: XnViewMP is finally a photo-editor for which you don’t have to pay. At the same time, it uses hardware very efficiently. In order to achieve more reasonable comparison times, we had to create an archive of up to 1024 photos, where we reduce the original resolution of 5472 × 3648 px to 1980 × 1280 px and filters with automatic contrast enhancement and noise reduction are also being applied during this process. We use 64-bit portable version 0.98.4.

Zoner Photo Studio X

Test environment: In Zoner Photo Studio X we convert 42 .CR2 (RAW Canon) photos to JPEG while keeping the original resolution (5472 × 3648 px) at the lowest possible compression, with the ZPS X profile ”high quality for archival”.

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

WinRAR 6.01

7-Zip 19.00

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

TrueCrypt 7.1a

Aida64 (AES, SHA3)

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Y-cruncher

---

Stockfish 13

Test environment: Host for the Stockfish 13 engine is a chess app Arena 2.0.1, build 2399.

Aida64, FPU tests

---

FSI (SPECworkstation 3.1)

Kirchhoff migration (SPECworkstation 3.1)

Python36 (SPECworkstation 3.1)

SRMP (SPECworkstation 3.1)

Octave (SPECworkstation 3.1)

FFTW (SPECworkstation 3.1)

Convolution (SPECworkstation 3.1)

CalculiX (SPECworkstation 3.1)

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

RodiniaLifeSci (SPECworkstation 3.1)

WPCcfd (SPECworkstation 3.1)

Poisson (SPECworkstation 3.1)

LAMMPS (SPECworkstation 3.1)

NAMD (SPECworkstation 3.1)

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Memory tests…

… and cache (L1, L2, L3)

---

---

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Processor power draw curve

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Average processor power draw

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Performance per watt

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Achieved CPU clock speed

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

CPU temperature/h2>

More aggressive power limits? Okay, but at a higher efficiency than Intel can do, AMD told itself while working on the Ryzen 7000 processors. And that’s how it really is, despite record-high clock speeds that are truly unusual for a new, unrefined manufacturing node. If this is just the beginning… In any case, not everything is rosy and AMD’s new processors have downsides that will need to be addressed in the future.

Conclusion

It seems that Intel will not have it easy at all with Raptor Lake. The Ryzen 9 7900X beats the Core i9-12900K by some 8–9 % in terms of maximum/multi-threaded performance (3D rendering, x264/x265 video encoding) with significantly lower power draw. Despite this, the increase in power draw of the R9 7900X over its predecessor (the R9 5900X) is very high – up to 40 %. To outperform the competing Ci9-12900K processor, it was necessary to stretch it this way. Only a better IPC (while maintaining AMD’s power limit at 142 W) would not be enough, and moving the PPT to 230 W is natural when the Zen 4 architecture is “ready” for such an extra load. So the power draw of the R9 7900X has gone up significantly compared to the R9 5900X, but importantly, the computing power has gone up in the same proportion as well, so the power efficiency does not practically decrease due to a significant increase in clock speeds. The difference in efficiency compared to the Core i9-12900K is (to the disadvantage of the processor) really significant.

However, AMD isn’t playing it for brute force at high load for just the entire processor, and it’s pushing the bar for single-threaded performance as well. It also determines the speed of responses in a normal office environment or on the web. This can also be seen in the top-notch results in PCMark’s practical tests, where the R9 7900X outperforms the Ci9-12900K, often by quite significant margins. A typical single-threaded task that wrings the most out of a single core is, for example, the encoding of an audio recording. During FLAC encoding, where we also track power consumption, the R9 7900X is 8 % faster and at the same time 10 % more efficient than the Core i9-12900K.

Ryzen 9 7900X maintains its high efficiency even in single-threaded tasks, and the extremely high clock speeds of over 5.6 GHz don’t change that either. Compared to the Ci9-12900K, the new Ryzen 9 (7900X) has better efficiency even in games, but overall it’s rather below average in this environment, even though gaming performance per watt is still 5 % higher than the R9 5900X. But compared to the R7 58003D with its large 3D V-cache cache and conservative clock speeds, the efficiency of the R9 7900X is a third weaker. It should be noted, however, that gaming performance is usually on the side of the Ryzen 9 “Raphael”.

There are exceptions where larger cache outweighs higher clock speeds (for example in Cyberpunk 2077, in F1 2020, in Shadow of the Tomb Raider or, most prominently, in Microsoft Flight Simulator), so on average, the R9 7900X is theoretically (at 720p@low, without the graphics card’s performance input) 5 % above the R7 5800X3D. At higher resolutions (1080p) at higher graphical details, when the CPU is already burdened with computing “call draws” the R9 7900X still has an edge over the R7 5800X3D, but it is “only” at the level of the Ci9-12900K. And rather just below it (by 1 %). This creates a very fertile ground for cherry-picking and presenting material to try and prove, based on specific titles and specific scenes, that this or that processor is significantly more powerful. In some games this may indeed be the case. Ryzen 9 7900X doesn’t “do too well” in Assassin’s Creed: Valhalla, in F1 2020 and even in Cyberpunk 2077, where at 1080p@high AMD’s processor most noticeably trails rival Intel. Still, in titles such as Shadow of the Tomb Raider, CS:GO or Total War Saga: Troy, the R9 7900X catches up and overall equalizes to a draw. With the difference that Ryzen 9 has lower power draw.

Also of note here is the broadcasting test, which we don’t point to all that often in the final summary.
However, it always refers to how much of the gaming performance the processors lose when capturing video using the x264 encoder. While the R9 7900X’s average fps loss is similar to that of the Ci9-12900K (±2 %), the significantly smaller drop in minimum fps (by 50 %) speaks in favour of AMD’s processor. This is both in OBS and in Xsplit.

The R9 7900X is also faster than the Ci9-12900K for video editing in DaVinci (in Adobe Premiere Pro, Ryzen often has slower Live Playback) as well as for Adobe After Effects graphics effects. The AMD processor is also in charge of the photo editing performance comparison. In Adobe Photoshop, the R9 7900X ends up with a better result 15 out of 18 filters. When exporting RAWs and generating thumbnails in Lightroom, the Ci9-12900K is again clearly slower (by 19–23 %). The situation doesn’t change in Zoner Photo Studio X either, and check out the tremendous difference in Topaz Labs AI applications, where batch edit times are shorter by almost half. But XnView is already better optimized for Core i9.

The biggest competitor from Intel (12900K) also loses to the R7900X in de/compression (~26 %) or in de/encryption (~42 %) and usually lags behind in numerical calculations, as well as in physical simulations. Anyway, how the R9 7900X stacks up against the Ci9-12900K is slowly becoming irrelevant. The Alder Lake generation is coming to the end of its lifecycle, and what matters is in what light AMD Raphael processors will appear against Intel Raptor Lake. However, AMD started this game well, except for the poorly thought-out cooling and high temperatures (but these have also been a problem for Intel processors for a long time). Even with high-performance air coolers such as the Noctua NH-U14S, it takes a perfectly functioning system cooling system to keep temperatures below 95 degrees Celsius (what AMD refers to as the upper limit of optimal temperatures).

When someone says to just slap a liquid cooler on it, don’t take their advice. The smaller ones, with a 240 mm radiator, often fall short of the performance of the more effective tower coolers, and when they are more powerful, it’s usually at the cost of significantly higher noise. Rather than an extreme cooler, AMD will have to figure out how to efficiently dissipate more heat from those increasingly smaller chiplets than the cooling interface now allows. However, we won’t put higher temperatures in the minuses, just as we don’t put it in the minuses for Intel processors either. It’s annoying, but it’s not overheating – these processors keep clock speeds and performance stable even at such high temperatures. Of course, the question is what effect such intense temperature has on durability. But we will not resolve that now.

After cross-sectional testing of all types, we can conclude that at least until the Raptor Lake generation is released, Intel has nothing to lay on the table against the Ryzen 9 7900X. Compared to the Core i9-12900K, AMD’s twelve-core processor overwhelmingly delivers higher performance (often quite significantly higher) at lower power draw. The overall platform’s viability will naturally also depend on what prices AM5 motherboards settle on. But regardless of them, the AMD Ryzen 7900X processor deserves the “Top-notch” editorial award due to its features.

English translation and edit by Jozef Dudáš

Testing games are from Jama levova

Special thanks also to Blackmagic Design (for DaVinci Resolve Studio license), Topaz Labs (for licenses to DeNoise AI, Gigapixel AI and Sharpen AI) and Zoner (for Photo Studio X license)

• Contents
• AMD Ryzen 9 7900X in detail
• Methodology: performance tests
• Methodology: how we measure power draw
• Methodology: temperature and clock speed tests
• Test setup
• 3DMark
• Assassin’s Creed: Valhalla
• Borderlands 3
• Counter-Strike: GO
• Cyberpunk 2077
• DOOM Eternal
• F1 2020
• Metro Exodus
• Microsoft Flight Simulator
• Shadow of the Tomb Raider
• Total War Saga: Troy
• Overall gaming performance
• Gaming performance per euro
• PCMark and Geekbench
• Web performance
• 3D rendering: Cinebench, Blender, ...
• Video 1/2: Adobe Premiere Pro
• Video 2/2: DaVinci Resolve Studio
• Graphics effects: Adobe After Effects
• Video encoding
• Audio encoding
• Broadcasting (OBS and Xsplit)
• Photos 1/2: Adobe Photoshop and Lightroom
• Photos 2/2: Affinity Photo, Topaz Labs AI Apps, ZPS X, ...
• (De)compression
• (De)cryption
• Numerical computing
• Simulations
• Memory and cache tests
• Processor power draw curve
• Average processor power draw
• Performance per watt
• Achieved CPU clock speed
• CPU temperature
• Conclusion