![AMD Ryzen 5 3600XT (AM4, 3.80 GHz, 6 -Core)](/im/Files/3/5/9/0/0/5/1/7/19257929-A_3rdGen_RYZEN5_3D_ROW_RF.png?impolicy=product&resizeWidth=720)
![](/im/Files/3/6/1/7/4/6/6/9/teaser21.jpeg?impolicy=teaser&resizeWidth=700&resizeHeight=350)
AMD Ryzen 5 3600XT: Better than its predecessor and the competition - and pointless
AMD is taking a swipe at Intel and is launching a refresh of three Ryzen 3000 desktop CPUs with the XT models. Although AMD is catching up with Intel in gaming with the Ryzen 5 3600XT, I can't see the point of the Matisse refresh.
Exactly one year after the original release, AMD is giving three of its Ryzen 3000 processors a refresh. According to AMD, they are manufactured in an optimised 7 nm process, which results in a 100 MHz higher boost clock for the 3600XT and 3900XT. For the 3800XT it is even 200 MHz. The XT processors are not intended to replace the existing models, but to complement them. I grabbed the smallest model, the Ryzen 5 3600XT, and compared it with the recently released competitor from Intel, the i5-10600K.
The chip in detail
The manufacturer has not changed the number of cores, the TDP or the base clock. Here is the processor compared to the 3600X and the competition from Intel.
Processor | Cores / Threads | Base / maximum boost clock (GHz) | Cache Total (MB) | PCIe Lanes | TDP (Watt) | Price (as of 20/05/2020) | Price per thread (as of 20/05/2020) |
---|---|---|---|---|---|---|---|
Ryzen 5 3600XT | 6 / 12 | 3.8 / 4.5 | 35 MB | 16 PCIe 4.0 | 95 | 279 | 23.25 |
Ryzen 5 3600X | 6 / 12 | 3.8 / 4.4 | 35 MB | 16 PCIe 4.0 | 95 | 231 | 19.25 |
Core i5-10600K | 6 / 12 | 4.1 / 4.8 | 12 MB | 16 PCIe 3.0 | 125 | 288 | 24 |
Test method and setup
I used the following components for the review:
I activate XMP in the BIOS. Otherwise I leave everything on default - except for overclocking, I'll get to that later. I use the drivers supplied by Asus and Intel as well as Windows 10 version 1909.
The test method is based on our graphics card reviews. I have revised the method since its introduction. Here is an overview of the various benchmarks:
- Cinebench R20
- CPU-Z Benchmark
- Blender
- Handbrake
- 7-Zip
- Photoshop
- PCMark 10
- Fire Strike / Fire Strike Ultra
- Time Spy / Time Spy Extreme
- Games: "Assassin's Creed Odyssey", "Civilisation VI: Gathering Storm", "Deus Ex: Mankind Divided", "Far Cry 5", "For Honor", "Gears 5", "Red Dead Redemption 2", "Strange Brigade" and "Shadow of the Tomb Raider"
I run all benchmarks three times and take the best result. I use the highest possible default settings for the games.
Overclocking and Cinebench R20
I limit my overclocking to the Noctua cooler. I am aware that this means I cannot make any statements about the maximum overclocking potential, but I can say after the tests: The 3600XT is relatively overclockable for a Ryzen processor.
As a reference, I run the Cinebench R20 benchmark. With stock settings, the 3600XT achieves a single core score of 521 and a multi-core score of 3676. The processor runs at 4.2 GHz on all cores and gets up to 75° Celsius. For comparison: The i5-10600K achieves a single core score of 505 and a multi-core score of 3604.
![](/im/Files/3/6/1/6/8/4/6/9/Cinebench%20R20%20hoher%20ist%20besser%203.png?impolicy=resize&resizeWidth=430)
With the 3600XT I achieve an overclock of 4.6 GHz on all cores. So 400 MHz more than on stock settings. I set a voltage of 1.4 V for this. The 3600XT gets up to 93° Celsius. In Cinebench R20, I achieve a score of 4109, which is ten per cent higher than on stock settings. This is relatively high for a Ryzen processor, as they are not considered to be so overclocking-friendly.
CPU-Z
In the CPU-Z benchmark, the Ryzen 5 3600XT is ahead in the Multi Core Score. However, the i5-10600K is ahead in the single core score.
![](/im/Files/3/6/1/6/8/4/6/8/CPU-Z%20Benchmark%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
7-Zip
In the integrated benchmark of 7-Zip - I select the standard "Dictionary size" of 32 MB - the 3600XT achieves 61,763 instructions per second (MIPS). That's just under ten per cent more than the i5-10600K with 56,207 MIPS.
![](/im/Files/3/6/1/6/9/6/2/2/7-Zip%20hoher%20ist%20besser.png?impolicy=resize&resizeWidth=430)
Blender bmw27
In the Blender bmw27 benchmark, the 3600XT is a little faster than the i5-10600K. It is two seconds ahead of the Intel chip.
![](/im/Files/3/6/1/6/9/6/2/1/Blender%20bmw27%20Benchmark%20tiefer%20ist%20besser.png?impolicy=resize&resizeWidth=430)
Handbrake
The 3600XT encodes the 88-second, 645 MB 4K trailer of "The Dark Knight Rises" with Handbrake's "Fast 1080p30" presets in 63 seconds. Compared to the i5-10600K, this is around 10 per cent faster.
![](/im/Files/3/6/1/6/9/6/2/0/Handbrake%20tiefer%20ist%20besser.png?impolicy=resize&resizeWidth=430)
Photoshop
Different workloads are performed in the Photoshop benchmark from Puget Systems. You can find more detailed information here. At the end, the benchmark calculates a score based on a reference workstation. The Ryzen 5 3600XT is four points behind the Intel processor, which corresponds to half a per cent.
![](/im/Files/3/6/1/6/9/6/1/9/Puget%20Systems%20Photoshop%20Benchmark%20hoher%20ist%20besser%201.png?impolicy=resize&resizeWidth=430)
PCMark 10
The PCMark 10 benchmark tests various scenarios such as the loading time of apps, efficiency in spreadsheets, browsing or photo and video editing. It calculates a score from this. With its 7157 points, the 3600XT performs around four per cent better than the i5-10600K.
![](/im/Files/3/6/1/6/9/6/1/8/PCMark%2010%20hoher%20ist%20besser%201.png?impolicy=resize&resizeWidth=430)
Fire Strike, Fire Strike Ultra, Time Spy and Time Spy Ultra
The synthetic game benchmarks provide a first look at the performance in games. I have omitted the overall score, which is calculated from the results of the graphics card and CPU. This is because the GPU score is very inconsistent. I had differences of over 1000 points here. This distorts the result enormously.
![](/im/Files/3/6/1/6/8/4/6/2/Physics%20Score%20Fire%20Strike%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/6/1/Physics%20Score%20Fire%20Strike%20Ultra%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
The Ryzen 3600XT is ahead in both Fire Strike benchmarks and the i5-10600K in both Time Spy benchmarks. The Ryzen chip is around 2.5 per cent faster in Fire Strike and the Core chip around 7 per cent faster in Time Spy.
![](/im/Files/3/6/1/6/8/4/6/0/CPU%20Score%20Time%20Spy%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
The games
In addition to the average FPS, I also provide you with the frame time in percentiles for the games, namely 99 and 99.9.
The measured values of the percentiles are frame times classically measured in milliseconds. In other words, the time intervals from picture to picture or frame to frame. The percentile values have the task of ignoring isolated outliers. 99 percentile means that 99 per cent of all measured values are faster than the specified measured value. If a value in the graphic is 95 FPS, 99 per cent are running at a higher frame rate than 95 FPS. Exactly one per cent, on the other hand, runs slower than 95 FPS. The same applies to the 99.9 percentile. For better comparability, the result is converted from frametimes in milliseconds to the traditional FPS value.
![](/im/Files/3/6/1/6/8/4/5/8/Assassins%20Creed%20Odyssey%201080p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/7/Assassins%20Creed%20Odyssey%201440p%20DX11%20hoher%20ist%20besser%203.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/6/Assassins%20Creed%20Odyssey%202160p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/5/Civilization%20VI_%20Gathering%20Storm%201080p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/4/Civilization%20VI_%20Gathering%20Storm%201440p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/3/Civilization%20VI_%20Gathering%20Storm%202160p%20DX11%20hoher%20ist%20besser%203.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/2/Deus%20Ex_%20Mankind%20Divided%201080p%20DX11%20hoher%20ist%20besser%204.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/1/Deus%20Ex_%20Mankind%20Divided%201440p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/5/0/Deus%20Ex_%20Mankind%20Divided%202160p%20DX11%20hoher%20ist%20besser%203.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/9/Far%20Cry%205%201080p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/8/Far%20Cry%205%201440p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/7/Far%20Cry%205%202160p%20DX11%20hoher%20ist%20besser%203.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/6/For%20Honor%201080p%20DX11%20hoher%20ist%20besser%203.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/5/For%20Honor%201440p%20DX11%20hoher%20ist%20besser%204.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/4/For%20Honor%202160p%20DX11%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/3/Gears%205%201080p%20DX12%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/2/Gears%205%201440p%20DX12%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/1/Gears%205%202160p%20DX12%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/4/0/Red%20Dead%20Redemption%202%201080p%20Vulkan%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/9/Red%20Dead%20Redemption%202%201440p%20Vulkan%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/8/Red%20Dead%20Redemption%202%202160p%20Vulkan%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/7/Strange%20Brigade%201080p%20Vulkan%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/6/Strange%20Brigade%201440p%20Vulkan%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/5/Strange%20Brigade%202160p%20Vulkan%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/4/Shadow%20of%20the%20Tomb%20Raider%201080p%20DX12%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/3/Shadow%20of%20the%20Tomb%20Raider%201440p%20DX12%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/2/Shadow%20of%20the%20Tomb%20Raider%202160p%20DX12%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
Over all games, the Ryzen 3600XT is behind the i5-10600K. The difference varies depending on the game. "Strange Brigade", for example, seems to be optimised for Intel processors. They perform much better than the AMD chips. At 1080p, the i5-10600K is around 13 per cent faster. This difference remains roughly the same for all resolutions.
![](/im/Files/3/6/1/6/8/4/3/1/Durchschnitt%20aller%20Games%201080p%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/3/0/Durchschnitt%20aller%20Games%201440p%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/2/9/Durchschnitt%20aller%20Games%202160p%20hoher%20ist%20besser%202.png?impolicy=resize&resizeWidth=430)
If I take out the outlier game "Strange Brigade", the difference is a lot smaller. At 1080p, the Intel chip is only 4.5 per cent faster.
![](/im/Files/3/6/1/6/8/4/2/8/Durchschnitt%20aller%20Games%201080p%20ohne%20Strange%20Brigade.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/2/7/Durchschnitt%20aller%20Games%201440p%20ohne%20Strange%20Brigade.png?impolicy=resize&resizeWidth=430)
![](/im/Files/3/6/1/6/8/4/2/6/Durchschnitt%20aller%20Games%202160p%20ohne%20Strange%20Brigade.png?impolicy=resize&resizeWidth=430)
Conclusion: Better deal than the i5-10600K - but still not worth it
With the Ryzen 5 3600XT, AMD catches up with Intel, especially in terms of single-core performance. In productive tasks, the 3600XT is usually ahead of the i5-10600K. In games, the new AMD chip still has to admit defeat. However, at less than five per cent, the difference is smaller than before.
I think the 3600XT is the better deal compared to the i5-10600K. Firstly, it's slightly cheaper. Secondly, you need a new motherboard with socket LGA 1200 for the Intel chip. A B450 board will also do for the 3600XT. There are already good boards for well under 150 francs. The Z490 boards for the Intel chip will set you back far more than 150 francs. Cheaper alternatives to the 3600XT would be the 3600X or 3600, which are already available for 231 and 188 francs respectively and only offer slightly less performance than the XT version.
This is precisely the point why the Matisse refresh makes no sense to me on the whole. AMD continues to produce the 3600X and the 3600 alongside the Ryzen 5 3600XT. If the XT version were to replace the X version at the same price, that would be a great thing. As it is, the manufacturer is merely offering a slightly more powerful version of the chip for a 17 per cent higher retail price. However, the performance increase is never 17 per cent. In the CPU buyer's guide that I published at the beginning of June 2020, I named the 3600X and the 3600 the best all-rounders. The Ryzen 5 3600XT doesn't change this statement. In terms of price-performance ratio, I would advise you to go for a 3600X or even a 3600.
![User Avatar](/im/Files/4/3/4/6/0/3/0/5/Autorenbild.jpg?impolicy=avatar&resizeWidth=96)
![User Avatar](/im/Files/4/3/4/6/0/3/0/5/Autorenbild.jpg?impolicy=avatar&resizeWidth=80)
From big data to big brother, Cyborgs to Sci-Fi. All aspects of technology and society fascinate me.