AMD’s new Ryzen 5 5600X is a great 6 core processor, but can it’s fewer yet faster Zen 3 cores beat the 8 in the Ryzen 7 3700X?
CPU spec differences
The key difference between these CPUs is of course the core and the thread counts. Both have 32mb of L3 cache, but the 3700X has a little more L2 cache. The 5600X has higher base and boost clock speeds, and both are 65 watt parts. The 3700X is currently just $25 more than the 5600X, but I’ve seen the 3700x under $310, and the 5600X currently has availability issues close to launch.
Test PC specs
Both processors were tested in the same system. I’ve tested with 32gb of DDR4-3200 memory running in dual channel at CL14 and with MSI’s GeForce RTX 3090 Gaming X Trio graphics card to minimize bottlenecks. Both processors come with an adequate cooler in the box, but they’re different, so I’ve tested with my Fractal S36 AIO with Noctua NT-H2 thermal paste for comparable results.
I’ve tested both CPUs at stock, and with the following all core overclocks applied. I was able to push the 5600X 400MHz higher than the 3700X, granted the 3700X does have more cores to power. With that in mind we’ll first check out the differences in various applications, as well as power draw and thermals, followed by gaming tests afterwards, then finish up by comparing some performance per dollar metrics.
Starting out with Cinebench R20 we can see the 5600X was scoring 18% higher in single core due to the improvements associated with Zen 3. The 3700X was still 11% faster when it came to multicore though, however this honestly doesn’t seem like that much given it also has 33% more cores. Once both are overclocked, the 3700X was only 8% ahead in multicore now, as the overclock on the 5600X is able to go a bit further. I’ve also tested the older Cinebench R15 as a lot of people still use it so you can compare my results, the margins weren’t too different to R20 so let’s move on. I’ve tested Blender with the BMW and Classroom benchmarks. This test makes use of all cores, so it’s no surprise that the 3700X is completing all tasks faster with its higher core count, granted it was only around 15% faster in the longer classroom test, and a similar margin with both overclocked. The V-Ray benchmark is another core heavy rendering workload, however at stock the 3700X was only 8% faster than the 5600X, and with both overclocked the 3700X’s lead lowers to 6%, which I think is a seriously impressive result when we keep in mind that the 5600X has two less cores. The Corona benchmark also uses the processor to render out a scene, and again the differences between the two aren’t that big considering the 3700X’s higher core count. It was less than 9% faster at stock, and 8% faster with both overclocked. Handbrake was used to convert one of my 4K laptop review videos to 1080p. Again the extra cores from the 3700X allow it to do better, with an above average difference here but still nothing too large. The 3700X was completing the task 16% faster at stock, then 12% faster with both overclocked. Adobe Premiere was used to export one of my laptop review videos at 4K. I’ve tested with both VBR 1 pass, which should make better use of hardware acceleration compared to 2 pass. The differences in 1 pass were very small, the 3700X was only 4% faster at stock, though in the 2 pass test it’s 12% faster. I’ve also tested Adobe Premiere but with the Puget Systems benchmark tool, as this tests for more things like live playback rather than just raw export times. The 3700X was only scoring 8% higher at stock, and just 5% higher with both overclocked, so not too big of a difference given this is 6 versus 8 cores. Adobe Photoshop was also tested with the Puget Systems benchmark tool. This test tends to favour single core performance, and as we’ve seen Zen 3 is certainly delivering in that regard, so it was no surprise to see the 5600X dominating this test and reaching a 17% higher score than the 3700X at stock. I’ve used 7-Zip to test compression and decompression speeds. There was basically no real difference in compression, the 3700X actually backtracked a bit when overclocked putting it behind the 5600X, however the 3700X was further ahead in decompression, 9% faster at stock and 7% faster with the overclocks. VeraCrypt was used to test AES encryption and decryption speeds, and this test saw the biggest difference out of all applications in favour of the 3700X. At stock the 3700X was over 33% faster, but then with both overclocked this lead lowers to around 25% better, so still a big win but the better overclock on the 5600X allows it to gain some ground. Microsoft Excel was tested using the Hardware Unboxed large number crunch test, and then they were both about the same at stock, so it makes sense that the 5600X was then faster with both overclocked due to the superior overclocking result. Like other single core tests, the 5600X was doing better here in Geekbench single core too. The 5600X was scoring more than 20% higher than the 3700X at stock in single core, and nearly 24% better with both overclocked. The extra cores in the 3700X still give it the win for multicore though, allowing it to score 9% higher at stock.
Application differences - Stock
Here’s how these two processors compare in all applications tested at stock. The 3700X was ahead in most cases, however these are the multicore tests where the extra cores are being taken advantage of. The 5600X is doing significantly better in terms of single core tests, which can be seen towards the bottom of the graph.
Application differences - Overclocked
Once both chips are overclocked things swing over to the 5600X a bit, as I was able to overclock it better than the 3700X, and this is something I’ve found possible with the Zen 3 processors I’ve tested so far. It could of course just be silicon lottery, but so far I’ve been able to push Ryzen 5000 further than 3000.
Power draw, thermals, clock speeds
Here’s how total system power draw from the wall looks with the blender test running. At stock, the 3700X system was using about 13% more power, and this was allowing it to complete the test 15% faster. With both overclocked though, the 5600X was consuming more power, though again I was able to overclock it a bit further, and this will vary based on the voltage you’re able to run the overclock at.
Higher power draw generally equals more heat, and the temperatures closely matched what we just saw in the power draw graph previously. At stock the 3700X was running a little warmer, but the 5600x temperatures really shot up with my overclock, not too bad but it was now 10 degrees Celsius warmer than the 3700X despite the 3700X still completing the task faster with its 8 cores. These are the clockspeeds being reached during these same tests, so at stock the 5600X is able to reach higher speeds. This combined with numerous other improvements to IPC is why the 5600X does better in single core, and when overclocked the 5600X went a little further.
Let’s get into the gaming results next, I’ve compared both processors in 9 games at 1080p, 1440p, and 4K resolutions. The 1080p results will of course be most useful for a CPU comparison, I’ve just included the higher resolutions so you can get an idea of how little the processor starts to matter.
Shadow of the Tomb Raider was tested with the games benchmark tool. The improvement seen at 1080p with the 5600X was one of the best out of all titles tested, with an average frame rate 30% higher than the 3700X. At 1440p the 5600X is still 15% faster, but then at 4K there’s no difference between them, same result.
Microsoft Flight Simulator was tested in the Sydney landing challenge. The 5600X had the biggest gains out of all games tested in this one at both 1080p and 1440p resolutions. Even its 1% lows were ahead of the averages the 3700X could offer, but again at 4K things change and there’s really not a noticeable difference. I’ve said this before, but Zen3 is no match for Ubisoft’s optimizations, or lack thereof.
Assassin’s Creed Odyssey saw one of the smallest differences regardless of resolution. The 5600X was ahead for the most part, but it’s only by a couple of frames, not a big deal. Death Stranding was tested by running through the same part of the game on both systems, because that’s all you do for the whole game. Like the others, basically the same performance at 4K though a little better 1% low from the 5600X, then at 1440p the 5600X was 9% faster than the 3700X in average FPS, and a larger 22% ahead at 1080p. Battlefield V was tested in campaign mode, and things were much closer together here, however both were basically hitting the 200 FPS frame cap at 1080p, no major differences regardless of the resolution in use. AMD claimed some big gains with Zen3 in CS:GO so I had to put that to the test, and sure enough the 5600X was reaching 24% higher average FPS at 1080p, 13% higher at 1440p, and even at 4K it was still 8% ahead, though the differences in 1% low performance were much smaller comparatively.
Red Dead Redemption 2 was tested with the games benchmark. Once more like most of the other titles, the 5600X had a nice boost over the 3700X at lower resolutions, with a 21% higher average frame rate at 1080p, and 13% higher at 1440p, but then 4K was just a couple of frames different.
Rainbow Six Siege was tested using the games benchmark with Vulkan. Again basically no difference in average frame rate at 4K, however the 1% low from the 5600X was 22% faster than the 3700X, so a more stable experience. Similar deal at 1440p, but then at 1080p the 5600X had a 13% higher average frame rate, with larger 26% boost seen to 1% low.Far Cry New Dawn was also tested with the games benchmark and tends to depend more on the processor than some others from my testing, which explains why even at 4K the 5600X is still 6% higher in average FPS, the second biggest difference at this resolution only behind CS:GO.
Over all 9 games tested, the 5600X was around 20% faster than the 3700X in average FPS, though this may be held back a little by the 200 FPS frame cap of battlefield V, which is why it’s down the bottom. In any case, the 5600X was definitely able to offer some nice gains.
Stepping up to 1440p, the 5600X was still 10% faster than the 3700X on average, though results could vary between no change to 20% depending on the specific game. Some titles are seeing pretty nice improvements even at this higher resolution with the 5600X.At 4K the 5600X is now less than 3% faster than the 3700X on average. The 3700X was never ahead, but the differences at this higher resolution just aren’t that major, and this is because the GPU is able to get to work making the processor selection matter less. As the 5600X is both cheaper and outperforms the 3700X in all games tested, it’s no surprise that it’s offering better value in terms of cost per frame across all three resolutions tested, though as mentioned earlier buying the 5600X at the moment close to launch for this price could prove difficult until supply issues are resolved.
Conclusion - 5600X or 3700X?
All things considered, I think the Ryzen 5 5600X offers serious gains over what AMD had available last generation. It blows the 3700X away in games despite having fewer cores, as 6 cores that are still fast is plenty for gaming. It could be interesting to do this comparison again a few years from now though, when perhaps those extra cores might make a bigger difference, but today if gaming is the focus, the 5600X just makes more sense. It’s cheaper and performs better. If you already have the 3700X it could be worth upgrading for games, especially if you’ve already got a motherboard that can take Zen 3.
Outside of gaming, the 5600X wasn’t too far behind the 3700X in multicore workloads. Improvements in other areas just allow it to make up significant ground over the 3700X despite having 2 fewer cores. If you really do need the best multicore performance though, then the 3700X does still make more sense there, but outside of that, the 5600X was much better in single core performance, and for less money, overall I think the 5600X is the way to go for most people who aren’t smashing all cores all the time. I’ve also compared the 5600X against the 3600 in this video over here, it’s definitely worth checking out if you want to see how Zen 2 compares with Zen 3 when both processors have the same core and thread count.