UserBenchmark: CPU Speed Test Tool
Speed test your CPU in less than a minute.
User Guide
Free Download
YouTube
Welcome to our freeware PC speed test tool. UserBenchmark will test your PC and compare the results to other users with the same components. You can quickly size up your PC, identify hardware problems and explore the best upgrades.
UserBenchmark of the month
Gaming
Desktop
ProGaming
CPUGPUSSDHDDRAMUSB
How it works
- — Download and run UserBenchmark.
- — CPU tests include: integer, floating and string.
- — GPU tests include: six 3D game simulations.
- — Drive tests include: read, write, sustained write and mixed IO.
- — RAM tests include: single/multi core bandwidth and latency.
- — SkillBench (space shooter) tests user input accuracy.
- — Reports are generated and presented on userbenchmark.
com.
- — Identify the strongest components in your PC.
- — See speed test results from other users.
- — Compare your components to the current market leaders.
- — Explore your best upgrade options with a virtual PC build.
- — Compare your in-game FPS to other users with your hardware.
Screenshots
Download free portable zip
|
PC 56,054,776 CPUs tested. |
CPU Score |
PC Score |
PC Status |
|---|---|---|---|
| 4 secs agoIntel Core i7-8700 | 88% | Nuclear submarine | 56% |
| 11 secs agoIntel Core i9-12900H | 103% | UFO | 57% |
| 15 secs agoIntel Core i5-4460 | 67% | Battleship | 62% |
| 16 secs agoIntel Core i7-1185G7 | 51% | Destroyer | 16% |
| 16 secs agoIntel Core i5-4690K | 75% | Aircraft carrier | 42% |
| 17 secs agoAMD Ryzen 5 5600X | 94% | Nuclear submarine | 20% |
| 18 secs agoAMD Ryzen 7 5825U Graphics | 85% | Aircraft carrier | 52% |
| 18 secs agoIntel Core i7-3770 | 67% | Battleship | 40% |
| 20 secs agoIntel Core i7-12700 | 113% | UFO | 80% |
| 22 secs agoIntel Core i7-9700K | 102% | Nuclear submarine | 29% |
| 29 secs agoIntel Core i5-6600K | 73% | Aircraft carrier | 31% |
| 33 secs agoIntel Core i9-9900K | 99% | UFO | 50% |
| 35 secs agoAMD Ryzen 5 2600 | 76% | Aircraft carrier | 76% |
| 36 secs agoAMD Ryzen 7 3700X | 77% | Battleship | 16% |
| 40 secs agoIntel Core i5-10600K | 90% | Nuclear submarine | 28% |
| 40 secs agoAMD Ryzen 7 3700X | 80% | Tree trunk | 32% |
| 49 secs agoAMD Ryzen 3 1300X | 61% | Battleship | 58% |
| 54 secs agoIntel Core i7-11700K | 107% | UFO | 77% |
| 56 secs agoIntel Core i5-12600K | 112% | UFO | 67% |
| 57 secs agoIntel Core i9-10850K | 103% | UFO | 63% |
| 58 secs agoIntel Core i7-5820K | 78% | Aircraft carrier | 57% |
| 1 min agoAMD Ryzen 5 2600 | 80% | Aircraft carrier | 85% |
| 1 min agoAMD Ryzen 9 3900XT | 93% | Nuclear submarine | 69% |
| 1 min agoAMD Ryzen 9 5900X | 105% | UFO | 59% |
| 1 min agoIntel Xeon E5-1620 v2 | 70% | Battleship | 47% |
| 1 min agoIntel Core i7-8700K | 91% | Nuclear submarine | 49% |
| 1 min agoIntel Core i5-2500K | 83% | Battleship | 45% |
| 1 min agoIntel Core i9-9900KF | 97% | Tree trunk | 60% |
| 1 min agoAMD Ryzen 7 3800X | 90% | Tree trunk | 56% |
| 1 min agoIntel Core i5-11300H | 69% | Battleship | 42% |
| 1 min agoIntel Core i7-8700K | 92% | Nuclear submarine | 56% |
| 1 min agoIntel Xeon E5-2697 v3 | 73% | Aircraft carrier | 25% |
| 1 min agoAMD Ryzen 7 3750H | 48% | Gunboat | 27% |
| 1 min agoIntel Core i5-7600K | 70% | Aircraft carrier | 50% |
| 1 min agoIntel Core i5-11400H | 84% | Nuclear submarine | 62% |
| 1 min agoAMD Ryzen 9 3900X | 88% | Nuclear submarine | 36% |
| 1 min agoAMD Ryzen 5 3600 | 92% | Nuclear submarine | 88% |
| 1 min agoAMD Ryzen 7 3700X | 89% | Aircraft carrier | 55% |
| 1 min agoAMD A10-5745M APU | 29% | Sail boat | 29% |
| 1 min agoIntel Core i5-11600KF | 100% | UFO | 67% |
| 1 min agoIntel Pentium G645 | 43% | Jet ski | 20% |
| 1 min agoIntel Core i7-12700KF | 124% | Tree trunk | 72% |
| 2 mins agoAMD Ryzen 9 5900X | 110% | Tree trunk | 75% |
| 2 mins agoAMD Ryzen 7 2700X | 89% | Nuclear submarine | 84% |
| 2 mins agoIntel Core i9-10900 | 101% | UFO | 62% |
| 2 mins agoIntel Core i7-7700K | 87% | Nuclear submarine | 59% |
| 2 mins agoAMD Ryzen 5 3600X | 85% | Aircraft carrier | 52% |
| 2 mins agoAMD FX-8320 | 60% | Battle cruiser | 28% |
| 2 mins agoIntel Core i7-10700K | 88% | Tree trunk | 32% |
| 2 mins agoIntel Core i7-4702HQ | 60% | Yacht | 46% |
About • User Guide • FAQs • Email • Privacy • Developer • YouTube
Feedback
Benchmarks — CG Director
Picking hardware can be tricky.
If you’re wondering what sort of hardware you’ll need for your workloads, you’ve come to the right place!
Benchmarks offer a sneak preview of performance – making them a reliable source of information that you can use before you decide on hardware. For example, a particular CPU workload might benefit from higher core counts while another might benefit from better clock speeds.
CG Director covers a variety of benchmarks to make your PC building experience simpler.
Pick the right hardware for your workload – ensuring that you get excellent value out of your brand-new workstation.
Need Help? We reply to every article comment! 🙂
CPUs with the highest Single-Core Performance [Updated List]
August 19, 2022
CPUs with high single core performance are ready-made for a variety of workloads you’re likely to run. PC Gaming, General Productivity, or Physical Simulations all greatly benefit from high single-core performance.
Intel CPU (Processor) List in Order of Performance
August 22, 2022
There are a myriad of Intel CPUs available to buy and it is difficult to compare them performance-wise. Here’s our Intel CPU List in order of performance.
AMD CPU (Processor) List in Order of Performance
August 29, 2022
Our AMD Processor List in Order of Performance will help you decide on the right AMD CPU for your needs. Multi-, Single-Core Performance and Value compared.
Cinebench R23 Scores [Updated Results]
August 30, 2022
Cinebench R23 is the newest instalment of the most popular CPU-Rendering Benchmark Cinebench. We have the Score Results for all modern Processors.
AMD Graphics Cards List in Order of Performance
May 11, 2022
A comprehensive List of all modern AMD Graphics Cards in Order of Performance in different Benchmarks such as Gaming and Rendering.
Corona Renderer Benchmark Results [Updated]
January 10, 2022
The Corona Benchmark is a popular CPU-Performance testing tool which makes uses of the popular CPU-Render Engine Corona.
CPU-Scores here.
Cinebench R20 Scores (Updated Results)
April 29, 2022
Regularly updated Cinebench R20 Scores List, sortable by multi-core and single-core score with in-depth info on CPU rankings.
VRAY Benchmark & CPU + GPU Scores (Updated Results)
March 31, 2022
Up-to-date V-Ray CPU and V-Ray GPU Benchmark list of all modern Graphics Cards (GPUs) & Processors (CPUs). Find out how well your hardware scores in the Vray Rendering Benchmark.
Redshift Benchmark Results (Updated)
June 4, 2022
Extensive Redshift Benchmark Results List with all modern GPUs, Operating Systems, CPUs and Multi-GPU-Setups. Find the best performing GPU for Redshift.
Cinebench R15 Scores (Updated Results)
November 28, 2021
An up-to-date comprehensive Cinebench Scores list of all modern Processors. Single and Multi-Core Benchmark. Compare the Cinebench Results here.
Blender Benchmark Results (Updated Scores)
April 23, 2022
The popular Open Source 3D Animation Software Blender has an in-built render-engine which can be used for testing CPU and GPU performance.
Best CPU & GPU Render Benchmarks
April 28, 2022
In-Depth Guide on how to use the best CPU and GPU Render Benchmarks to test your System’s performance, make your hardware perform optimally and solve possible performance problems.
Best CPU For Rendering [2022 Guide]
May 9, 2022
Processor (CPU) Performance / Price Ratio for 3D Rendering. Ultimate List for finding the best CPU for Rendering.
TOP 5 benchmarks for testing the processor
CPU Benchmark (central processing unit performance test) is a special software designed to check the stability of the CPU and to compare the obtained performance results with all current processor models.
Most modern PC users run a CPU performance test only when they buy a computer. Yes, this is certainly correct if you check how the computer you buy meets expectations and how stable it works. This is especially true if the processor or system unit is bought not in a store with a guarantee, but from the hands of an ad.
But besides this, it is useful to drive the processor with a benchmark much later after the purchase.
After all, the test will show not only the cherished «parrots», but will also allow you to see how the operating and extreme temperatures of the CPU correspond to the recommended ones. After all, over time, the processor fan may stop giving a sufficient number of revolutions to cool the heatsink, and the thermal paste between the heatsink and the processor may dry out and deform. Which will eventually cause the processor to overheat and make it inoperable. Notagram.ru has specially selected the best free benchmarks for testing the CPU.
TOP-5 benchmarks for testing the processor
Cinebench
Photo: Cinebench / MAXON Computer GmbH
Cinebench is one of the benchmarks for testing and determining the level of processor performance and stability of their work. This software is specially «charged» to maximize the load on multi-core CPUs, as well as the video system.
That in general will allow you to understand how the computer can seamlessly cope with peak multimedia loads. So if you’re buying a PC for gaming, streaming, and multimedia, always use Cinebench to evaluate stability and performance.
download Cinebench
AIDA64
Photo: AIDA64/FinalWire Ltd.
AIDA64 is the most famous, simple and intuitive software for testing and identifying computer components. Everything is as clear as possible. Even a beginner can easily figure out the settings and functionality. I selected the processor icon in the menu, noted what and how to check, and launched the test. You can check both the stability of the CPU, and get data on its performance relative to others. The only caveat is that the program can be used for free only for the first thirty days. But this is quite enough to check the CPU, for example, when buying a computer.
download AIDA64
OCCT
Photo: OCCT/OCCT
OCCT (OverClock Checking Tool) is one of the best tools for testing the computer’s CPU.
This software offers three different types of CPU tests, which clearly demonstrate all the important parameters of the processor operation on graphs and monitoring indicators. From its heating on a specific core, to operating frequencies and voltage. At first, OCCT may seem too complicated to understand all its settings, but there really is nothing to worry about. Everything is quite simple and intuitive.
download OCCT
LinX
Photo: LinX/Dualist
LinX is another development from our craftsmen. This benchmark allows you to maximize the load on the computer’s CPU in conjunction with RAM, and check how accurately it performs complex mathematical calculations. Despite the ascetic design and interface, this program allows you to conduct stress testing as accurately and correctly as possible. Included in any first aid kit of a real sysadmin and iron worker.
download LinX
Prime95
Photo: Prime95/George Woltman
Prime95 is one of the earliest and well-known tools for testing personal computer processors.
This software allows you to determine both the performance of your CPU and run it through a stress test. Choose the «Benchmark» or «Torture Test» mode and the processor will be loaded to the very eyeballs by searching for Mersenne primes. Many people use Prime95 in conjunction with some kind of monitoring software. Since Prime9 himself5 does not have visual CPU monitoring tools.
download Prime95
Preview photo: Christian Wiediger/Unsplash
If you find a mistake, please highlight the text and press Ctrl+Enter .
Almost forgot, more interesting
Testing seven processors using the updated iXBT Application Benchmark 2018
This article can be considered as an addition to the announcement of our new iXBT Application Benchmark 2018 test package, where all aspects of the new testing methodology were considered in detail, as well as new reference results were announced.
Recall that the reference system in the iXBT Application Benchmark 2018 test package uses a computer with a six-core Intel Core i7-8700K processor of the following configuration:
| Processor | Intel Core i7-8700K |
|---|---|
| Motherboard | Asus Maximus X Hero |
| Chipset | Intel Z370 Express |
| Memory | 16 GB DDR4-2400 (dual channel) |
| Graphics | Intel UHD Graphics 630 |
| Accumulator | SSD Seagate ST480FN0021 480 GB (SATA) |
| Operating system | Windows 10 Pro (64-bit) |
To calculate the integral performance score, initially the results of all tests are normalized relative to the results of the reference system.
The normalized result is obtained by dividing the task execution time by the reference system by the task execution time by the system under test. Next, the normalized results of all tests are divided into eight logical groups, and for each group of tests, its own integral result is calculated as the geometric mean of the normalized results. For the convenience of presenting the results, the resulting value is multiplied by 100.
Our test suite has a logical group of tests called File Operation Speed. This group includes two tests, and their results are determined primarily by the performance of the drive. The results of all other tests, which are divided into seven logical groups, are mainly determined by the performance of the processor. They also depend (albeit to a lesser extent) on the performance of the graphics subsystem and memory, but almost do not depend on the performance of the drive. That is why for these seven logical groups of tests a separate integral result is calculated as the geometric mean of the intermediate integral results for the seven groups of tests.
We call this integral result the platform result.
The final integral result is defined as the geometric weighted average of the integral result of the platform tests and the integral result of the drive tests. For the integral result of the drive tests, the weight coefficient is 0.3, and for the integral result of the platform tests — 0.7.
Test Bench Configuration
In this article, we will review the results of testing seven processors using our new methodology. The processors themselves are by no means new, we have already tested all of them, so we will leave their detailed technical characteristics and power consumption out of consideration. The point of testing these processors is to make sure that the new benchmark adequately evaluates the performance of various systems — the results of a reference system alone are not enough for this. In addition, the reference system does not have discrete graphics (only the processor graphics core), and I would like to understand how the presence of powerful discrete graphics can affect the test results.
We ended up adding seven more different configurations to our reference system: two based on the Coffee Lake family of processors and five based on the Skylake-X family of processors.
To test the Skylake-X family processors, the following configuration was used:
| Motherboard | Asus TUF X299 Mark I (BIOS version 1102) |
|---|---|
| Chipset | Intel X299 |
| Memory | 16 GB DDR4-2400 (quad-channel) |
| Graphics | MSI GeForce GTX 1070 Gaming X 8G |
| Accumulator | SSD Seagate ST480FN0021 480 GB (SATA) |
| Operating system | Windows 10 Pro (64-bit) |
It is worth recalling separately why we indicate the UEFI BIOS version of the motherboard. The thing is that initially the Asus TUF X29 board9 Mark I came to us with UEFI BIOS version 0802. We did not notice any problems, the board started with a 10-core Intel Core i9-7900X processor, and with a 16-core Intel Core i9-7960X, and with an 18-core nuclear Intel Core i9-7980XE.
Moreover, with the Core i9-7900X, the board passed the test quite successfully — all the tests of the iXBT Application Benchmark 2018 package. But problems began with the Intel Core i9-7960X and Intel Core i9-7980XE processors. The Core i9-7960X failed to pass any tests: the processor overheated and the system rebooted or simply hung. With Core i9-7980XE the situation was not much better. The problem, apparently, was that the processor was supplied with an overvoltage at slightly raised frequencies, and the cooling system could no longer cope. True, a rather simple air cooling system was used on the test bench, and for such processors you need a CBO that can remove at least 165 W of heat, but freezes and spontaneous reboots are still unacceptable. Performance should decrease, but everything should work stably.
After flashing the new UEFI BIOS version (version 1102), the situation has changed dramatically. The processors began to work stably, without overheating — however, not as «fast» as before.
Clock speeds have become a little lower. According to our estimates, for the Intel Core i9-7900X processor, the integral performance drop after updating the UEFI BIOS was 3%, which is not much.
To test the processors of the Coffee Lake family, the following configuration was used:
| Motherboard | Asus Maximus X Hero |
|---|---|
| Chipset | Intel Z370 Express |
| Memory | 16GB DDR4-2400 (dual channel) |
| Graphics | MSI GeForce GTX 1070 Gaming X 8G |
| Accumulator | SSD Seagate ST480FN0021 480 GB (SATA) |
| Operating system | Windows 10 Pro (64-bit) |
As you can see, both stands use the MSI GeForce GTX 1070 Gaming X 8G video card. For processors of the Skylake-X family, the presence of a video card is a prerequisite, since these processors do not have an integrated graphics core. The processors of the Coffee Lake family have a graphics core, but in order to equalize bench configurations as much as possible and be able to compare the processors, and at the same time see which test results depend on the performance of the graphics subsystem, we also used a video card in the bench for Coffee Lake processors MSI GeForce GTX 1070 Gaming X 8G.
Five models have been tested in the Skylake-X family:
| Core i7-7820X | Core i9-7900X | Core i9-7940X | Core i9-7960X | Core i9-7980XE | |
|---|---|---|---|---|---|
| Number of cores/threads | 8/16 | 10/20 | 14/28 | 16/32 | 18/36 |
| Base clock, GHz | 3.6 | 3.3 | 3.1 | 2.8 | 2.6 |
| Maximum clock frequency in Turbo Boost mode, GHz | 4.3 | 4.3 | 4.3 | 4.2 | 4.3 |
| Turbo Boost Max 3.0 GHz | 4.5 | 4.5 | 4.4 | 4.4 | 4.4 |
| L3 cache, MB | 11 | 13.75 | 19.25 | 22 | 24.75 |
| TDP, W | 140 | 140 | 165 | 165 | 165 |
| Number of PCIe 3.0 lanes | 28 | 44 | 44 | 44 | 44 |
Two models were tested in the Coffee Lake processor family:
| Core i7-8700K | Core i5-8400 | |
|---|---|---|
| Number of cores/threads | 6/12 | 6/6 |
| Base clock, GHz | 3. 7 |
2.8 |
| Maximum clock frequency in Turbo Boost mode, GHz | 4.7 | 4.0 |
| Graphics core | Intel UHD 630 | Intel UHD 630 |
| L3 cache, MB | 12 | 9 |
| TDP, W | 95 | 65 |
| Number of PCIe 3.0 lanes | 16 | 16 |
The Core i7-8700K processor test bench differs from our reference system only in the presence of the MSI GeForce GTX 1070 Gaming X 8G graphics card, so comparing the results of the reference system and the system with the Core i7-8700K processor and the MSI GeForce GTX 1070 Gaming X graphics card 8G will allow us to appreciate the impact of a powerful graphics card.
Test results
We will present the test results in diagrams. The name of the tested configuration corresponds to the name of the processor model, and the calculation of the integral results is performed relative to the results of the reference system, for which the integral results in each logical group of tests are taken equal to 100 points.
In addition, we will also consider the test results normalized relative to the reference system in each individual application.
So, let’s start with the general integral result without taking into account the data storage subsystem (integral result of the platform).
Actually, everything is logical here, except that the result of the Core i9-7980XE processor is almost the same as that of the Core i9-7940X processor, and slightly inferior to the result of the Core i9-7960X processor. But there is a logical explanation for this. First, not all applications need the number of cores that Core i9 processors offer.-7940X and above, that is, not all tasks can be effectively parallelized by more than 28 threads. Secondly, the more cores the processor has, the lower the frequency of their work. And here you need to choose which is better: a large number of slow cores or a not very large number of faster cores. Based on base frequencies, the situation is as follows: the Core i9-7960X processor has 14% more cores than the Core i9-7940X processor, but their base clock speed is 9.
7% lower than the Core i9-7940X processor. In Core i9 processorThe -7980XE has 12.5% more cores than the Core i9-7960X, but its base clock speed is 7% slower than the Core i9-7960X. In reality, of course, the base frequency says little. The Core i9-7980XE processor operates at a frequency clearly below 2.6 GHz, and as a result, in terms of the integrated platform performance indicator, a system based on the top Core i9-7980XE processor is inferior to a system based on the Core i9-7960X processor.
In general, if we talk about top-end and, by the way, very expensive Core i9 processors-7940X, Core i9-7960X and Core i9-7980XE, the main problem is how to cool them. Air cooling clearly cannot cope with them, which means that effective water cooling is needed, capable of removing at least 165 watts of heat. Then, perhaps, the results of the Core i9-7980XE processor will be higher than that of the Core i9-7960X processor. However, we did not have such an effective CBO at the time of the tests.
Comparing the results of testing a system based on the Core i7-8700K processor with the results of testing a reference system, we can conclude that the integral performance of the platform depends not only on the performance of the processor, but also on the performance of the video card. We’ll talk about the specific tests in which the presence of a video card is important a little later.
The integral performance results of the entire system as a whole, that is, taking into account tests that depend on the drive, approximately repeat (but with different scores) the integral platform results, which is quite logical, since the integral results of tests that depend on the performance of the drive are almost the same for all systems under test (they use the same SSD).
Video conversion
Well, now let’s analyze the test results in a little more detail. And let’s start with the Video Converting logical group, which includes tests based on applications MediaCoder x64 0.
8.52.5920, HandBrake 1.07 and VidCoder 2.63. Recall that the VidCoder application is based on HandBrake and differs from it only in the interface, which in some cases is more convenient. The MediaCoder, HandBrake, and VidCoder application-based tests use different tasks.
MediaCoder converts a 4K video clip (3840×2160) in MPEG4 AVC format with a bit rate of 42.2 Mbps, which is converted using the H.264 (x264) codec with frame size reduction to 1920×1080 and downscaling video bitrate up to 16 MB/s.
HandBrake uses the same 4K video file as MediaCoder, but uses the H.265 MKV 1080p30 preset for conversion.
VidCoder uses the video conversion task of a 4.33GB DVD5 (MPEG-2, 25 FPS) movie as a test. The film is converted by the H.264 (x264) codec with the Normal preset.
So, as follows from the test results, in the video conversion tasks that we use in our tests, the result does not depend on the GPU, that is, for the integrated Intel UHD 630 graphics core in the processor, and for the powerful MSI GeForce GTX 1070 Gaming video card X 8G results in all three tests are the same.
In general, the best result in this case is shown by the Core i9-7960X processor, which is a kind of golden mean between the number of cores and their frequency at the level of cooling that we were able to provide.
Interestingly, the results for the Core i7-8700K, Core i9-7940X, Core i9-7960X, Core i9-7980XE processors are almost the same in the VidCoder application-based test.
Rendering
The Rendering group includes four tests based on applications POV-Ray 3.7, LuxRender 1.6 x64 OpenCL, Blender 2.79and Adobe Photoshop CC 2018. In Adobe Photoshop CC 2018, we render 3D text.
As follows from the test results, in the rendering tests, the result is almost independent of the GPU, that is, for the integrated Intel UHD 630 graphics core in the processor, and for the powerful MSI GeForce GTX 1070 Gaming X 8G video card, the results are almost the same. The only exception is the test based on the Adobe Photoshop application, in which the rendering speed is slightly higher when using a powerful video card.
In all other tests, the use of a powerful video card does not affect the result in any way.
As you know, rendering tasks are very well parallelized: for them, the more cores, the better. But this is only if the clock frequency remains unchanged. In reality, with an increase in the number of cores with a simultaneous decrease in their clock frequency, there may also be a loss in performance. In particular, the Core i9-7980XE processor shows a result even slightly lower than the Core i9-7960X processor.
Speaking specifically for applications, in POV-Ray, LuxRender and Blender, the Core i9 processor shows the best result-7960X, and in Adobe Photoshop — Core i9-7940X.
Video Content Creation
The Video Content Creation group includes five tests based on Adobe Premiere Pro CC 2018, Magix Vegas Pro 15, Magix Movie Edit Pro 2017 Premium v.16.01.25, Adobe After Effects CC 2018 and Photodex ProShow Producer 9.0.3782.
The test uses Adobe Premiere Pro to create a 4K (3840×2160) resolution MPEG4 AVC video clip from five separate video clips at 42.
2 Mbps. Stage rendering and video file export are performed using the HEVC (H.265) codec.
The Magix Vegas Pro test creates a four-segment video with the same specifications as the Adobe Premiere Pro CC 2018 test. The video is exported using the H.264 codec downsampled to Full HD. To do this, video rendering is used with the Internet HD 1080p 29.97 fps preset.
The Magix Movie Edit Pro test also creates a four-segment video with the same specifications as the previous two tests, but the project is exported using the HEVC (H.265) video codec with an MP4 UltraHD 3840×2160 preset 29.97 fps
In the test using the Adobe After Effects application, the final export of the project is made without compression, in the AVI format.
In Photodex ProShow Producer, the test task is a project created from 20 digital photos taken with a Canon EOS 5D Mark III camera and converted to JPEG format at a resolution of 5760×2840. When creating a movie, various transition effects are applied between individual slides, and the slides themselves are animated.
The final movie is created with MPEG-4 H.264 1080p (60 fps Extreme Quality) preset.
According to the test results, in the video content creation tests, the result is GPU-dependent, that is, the results for the MSI GeForce GTX 1070 Gaming X 8G graphics card are slightly higher than for the Intel UHD 630 processor graphics core. However, out of five tests In this group, only the test based on Adobe Premiere Pro demonstrates dependence on GPU performance.
According to the integral performance result in the Video Content Creation group, the situation is almost the same as in the previous case, that is, the Core i9 processor is in the lead-7960X, which is even slightly higher than the Core i9-7980XE.
In terms of applications, in Photodex ProShow Producer, Magix Movie Edit Pro and Adobe After Effects, the results for the Core i9-7900X, Core i9-7940X, Core i9-7960X and Core i9-7980XE processors are almost the same. It seems that in these tests, increasing the number of threads above 20 does nothing.
We also note that for the test based on the Photodex ProShow Producer application, the best result is achieved on the Core i7-8700K processor (12 threads and high clock speed).
Digital Photo Processing
The Digital Photo Processing group includes three tests based on Adobe Photoshop CC 2018, Adobe Photoshop Lightroom Classic CC 2018 and Phase One Capture One Pro v.10.2.0.74. All tests use batch processing of 100 RAW photos taken with a Canon EOS 5D Mark III. The resolution of each photo is 3840×5760.
In the test based on Adobe Photoshop, processing each photo includes the following sequence of actions: opening a RAW file, applying a Lens Correction filter, applying a Reduce Noise filter, reducing the photo to 800 pixels in height, and saving as a JPEG with maximum quality .
In the test based on the Adobe Photoshop Lightroom application, a catalog is first created, then one (first) frame of the catalog is processed, after which the settings are synchronized, and at the final stage the catalog is exported with the settings synchronized by the first frame.
The directory is created without copying. The processing of one single frame in this case consists in applying a lens correction filter and a noise reduction filter (Noise Reduction (Luminance 50% and Color 50%)). Export is made in JPEG format with default settings (quality 60%, resolution 240 dpi).
The Phase One Capture One Pro test initially imports a collection of photos into a project and creates a preview. Then batch processing of photos in the mode of automatic quality improvement (Auto Adjustment) is performed. The final step is to save the photos in JPEG format. When saving photos in JPEG format, the JPEG sRGB preset is used.
Perhaps the test results in the Digital Photo Processing group are the most interesting. Firstly, according to the integral result, the Core i7-8700K processor is in first place. Secondly, for all processors of the Core-X family, the result is approximately the same. And thirdly, comparing the results for a system based on the Core i7-8700K processor with the results of the reference system, we can conclude that the results are highly dependent on the GPU, that is, with the MSI GeForce GTX 1070 Gaming X 8G video card, the integrated result is significantly better than with Intel UHD 630 graphics processor.
And now specifically for applications.
In the test based on the Adobe Photoshop application, the Core i7-8700K processor is in first place, and for all processors of the Core-X family, the result is approximately the same and almost the same as for the Core i5-8500 processor. The results of the test based on the Adobe Photoshop application do not depend on the performance of the GPU.
In a test based on the Adobe Photoshop Lightroom application, the results for the Core i9-7980XE and Core i9-7960X processors are the same. The results of this test are also independent of GPU performance.
In the Phase One Capture One Pro test, the Core i7-8700K performs best and the Core i9-7980XE performs worst. In this test, the use of the MSI GeForce GTX 1070 Gaming X 8G graphics card in combination with the Intel Core i7-8700K processor allows you to increase the speed of the test by 3.5 times! That is, this test is more determined by the performance of the GPU than the CPU.
Text Recognition
The Text Recognition group includes only one test based on the Abbyy FineReader 14 OCR application, since de facto this application simply has no competitors.
A test based on the Abbyy FineReader application measures the recognition time of a multi-page PDF document.
As follows from the test results, in this test the result does not depend on the GPU in any way, that is, for the MSI GeForce GTX 1070 Gaming X 8G video card, the result is exactly the same as for the Intel UHD 630 processor graphics core.
By the integrated result performance in this test, everything is as it should be. The highest result (in points) is demonstrated by the top-end Core i9 processor-7980XE.
Archiving
The «Archiving» group includes two tests: based on the WinRAR 5.50 application and based on the 7-Zip 18 application. In both cases, a 9.15 GB folder is compressed, which contains various types of data (photos, videos, PDF documents, etc.
), but WinRAR and 7-Zip archivers use different compression settings.
As follows from the test results, in the archiving tests, the result does not depend on the GPU, that is, the results for the MSI GeForce GTX 1070 Gaming X 8G video card are exactly the same as for the Intel UHD 630 processor graphics core.
Core i9-7940X demonstrates the best result in the integrated performance result in the Archiving group. But this is precisely the integral result for the two tests, since the results differ in different tests.
More specifically, for the Core i5-8400 processor, the normalized speeds in tests based on WinRAR and 7-Zip applications are the same. Similarly, these speeds are the same for the Core i7-8700K, Core i7-7820X and Core i9-7900X processors, but naturally they get higher for the Core i7-8700K processor, even higher for the Core i7-7820X processor, and even higher for the Core processor. i9-7900X.
For the Core i9-7940X processor, the normalized speeds are even higher, but no longer equal to each other: in the test based on the 7-Zip application, the speed is higher than in the test based on the WinRAR application.
But in the case of the Core i9-7960X and Core i9-7980XE processors, miracles begin. In the test based on the 7-Zip application, the normalized speed continues to increase, but in the test based on the WinRAR application, the normalized speed drops sharply. As a result, the test based on the 7-Zip application is best performed by the Core i9 processor.-7980XE, and in the test based on the WinRAR application, the best result is demonstrated by the Core i9-7940X processor. What exactly is the reason for such a drop in speed in the test based on the WinRAR application when using Core i9-7960X and Core i9-7980XE processors, you need to understand separately. Within the framework of this article, we do not set such a task for ourselves, but in the future we will definitely return to this issue.
Scientific calculations
The Scientific calculations group includes four tests based on LAMMPS, NAMD 2.11, Mathworks Matlab R2017b and Dassault SolidWorks Premium Edition 2017 SP4.
2 with SolidWorks Flow Simulation 2017.
The LAMMPS molecular dynamics application is used to calculate interactions of up to tens of millions of atoms. The test uses the calculation of two problems that come with the LAMMPS package as an example.
The NAMD Molecular Dynamics test uses one specific project calculation.
In the test based on the application Mathworks Matlab R2017b, the solution of the problem of vibrations of a rectangular membrane with fixed edges is used as a test problem.
Test based on Dassault SolidWorks with SolidWorks Flow Simulation to determine the speed of calculations in fluid dynamics and heat transfer problems. The test uses a thermal calculation of a computer case that contains heat-producing elements and a single fan. This project is included in the Flow Simulation package as an example.
As follows from the test results, they do not depend on the performance of the graphics subsystem in any way, that is, these are purely processor tests.
According to the integral performance result in the «Scientific calculations» group, the best result is demonstrated by the Core i9-7960X processor, whose result is slightly higher than that of the Core i9-7980XE processor.
Now let’s look at the results for each test separately.
In a test based on the Dassault SolidWorks application, the Core i7-8700K performs best. That is, in this test, the clock frequency is more important than the number of processor cores.
The Mathworks Matlab test also does not benefit significantly from a large number of processor cores. For 10-core, 14-core, 16-core and 18-core processors, the results in the test based on this application differ little from each other and only slightly (no more than 20%) outperform the result of the Core i7-8700K processor.
LAMMPS and NAMD application-based test scores increase as the number of processor cores increases, with the exception of the 18-core Core i9-7980XE processor. In the LAMMPS application-based test, its result is the same as the Core i9-7960X processor, and in the NAMD application-based test, it is even lower than the Core i9-7960X processor.
File Operations
And the last group is «File Operations». This group includes two tests, the results of which are determined by the performance of the system drive. The first test is a simple copy of data from one location on the system drive to another location on the same drive. Copying is done using the Windows 10 operating system.
In the second test, using the WinRAR 5.50 archiver, an uncompressed archive is created, that is, packing many separate files into one file.
Naturally, there is practically no processor load in this group of tests, and the results are determined only by the performance of the drive and the I/O subsystem of the board. The I/O subsystem of the Intel X299 chipset is more sluggish than that of the Intel Z370 chipset (this has been tested many times), so the results of the Core i5-8400 and Core i7-8700K processors in these tests are the same and slightly better than for the Core-family processors. x.
Conclusion
In this article, we have covered seven processor models that we tested using our updated methodology using the iXBT Application Benchmark 2018.