Ryzen 7 2700X vs. Ryzen 7 3700X: High Refresh Gaming Comparison

So, kudos to the team for doing this work but none of these games are a problem for my CPU. Any CPU maxes my 144hz monitor in these titles.

I’d be far more interested in the games that my CPU struggles to hit 144 with. For example FFXV, it would be nice to know which part will get me closest to the 120hz cap and how much better it is than the alternatives. Same with Far Cry 5, AC Odyssey etc. I’d also be interested in seeing which CPU can complete turns the quickest in Civ 6.

Also from my experience, average results on CPU tests are quite misleading as the variance can be quite high. It would be good to see which CPUs suffer the most during the most intensive parts of the game.

But still, appreciate all the work you already do.
 
Based on just about every single benchmark test you have up there the choice is clear: Buy Intel.

The 10600k and 3700x are both just below $300 and within a $15 difference of each other.
For gaming, the 10600k performs similarly to the 9900k (the better overall CPU than both).
 
Based on just about every single benchmark test you have up there the choice is clear: Buy Intel.

The 10600k and 3700x are both just below $300 and within a $15 difference of each other.
For gaming, the 10600k performs similarly to the 9900k (the better overall CPU than both).

I disagree, it's faster if all you do is game, for testing esports it would have made more sense to load twitch in the background streaming at HD like alot of esports players these days.
 
Hi, any chance you would consider doing a CPU benchmark for a few older games? All the games you have benchmarked above enjoy frames rates well over 100 FPS. When I play Starcraft 2: Legacy of the Void, I get frame drops to 48 FPS in the introduction of Amon's Reach when there are not many units. I don't play multiplayer. I replayed Crysis Warhead recently and the frame rate is terrible on some levels, In All the Fury, my frame rate drops below 30 FPS and stutters a lot of the time. Both games at max settings, for Crysis, 4x anti-aliasing is enabled also, 2560 x 1440 resolution, i5 2500k, GTX 1660 Super. Thanks.
 
Last edited:
Hi, any chance you would consider doing a CPU benchmark for a few older games? All the games you have benchmarked above enjoy frames rates well over 100 FPS. When I play Starcraft 2: Legacy of the Void, I get frame drops to 48 FPS in the introduction of Amon's Reach when there are not many units. I don't play multiplayer. I replayed Crysis Warhead recently and the frame rate is terrible on some levels, In All the Fury, my frame rate drops below 30 FPS and stutters a lot of the time. Both games at max settings, for Crysis, 4x anti-aliasing is enabled also, 2560 x 1440 resolution, i5 2500k, GTX 1660 Super. Thanks.

Yeah I agree! It’s not just competitive gamers who want high refresh rate.
 
All these tests are on low quality but somehow "demanding"? Textures DO impact FPS quite a bit so I don't see how this is in any way representative of real-world usage.
 
Hi, any chance you would consider doing a CPU benchmark for a few older games? All the games you have benchmarked above enjoy frames rates well over 100 FPS. When I play Starcraft 2: Legacy of the Void, I get frame drops to 48 FPS in the introduction of Amon's Reach when there are not many units. I don't play multiplayer. I replayed Crysis Warhead recently and the frame rate is terrible on some levels, In All the Fury, my frame rate drops below 30 FPS and stutters a lot of the time. Both games at max settings, for Crysis, 4x anti-aliasing is enabled also, 2560 x 1440 resolution, i5 2500k, GTX 1660 Super. Thanks.

Starcraft 2 is optimized like a trash heap and it has frame drops no matter what CPU you use. It's not a good game to benchmark with.
 
Also from my experience, average results on CPU tests are quite misleading as the variance can be quite high. It would be good to see which CPUs suffer the most during the most intensive parts of the game.

But still, appreciate all the work you already do.

What you are describing is 1% and 0.1% lows. Of which this review already contains 1% lows.
 
Hi, any chance you would consider doing a CPU benchmark for a few older games? All the games you have benchmarked above enjoy frames rates well over 100 FPS. When I play Starcraft 2: Legacy of the Void, I get frame drops to 48 FPS in the introduction of Amon's Reach when there are not many units. I don't play multiplayer. I replayed Crysis Warhead recently and the frame rate is terrible on some levels, In All the Fury, my frame rate drops below 30 FPS and stutters a lot of the time. Both games at max settings, for Crysis, 4x anti-aliasing is enabled also, 2560 x 1440 resolution, i5 2500k, GTX 1660 Super. Thanks.

4c4t CPUs (especially older ones) are going to suffer in some dx11-dx9 titles, old and new. Try dxvk (vulkan emulator). With a similar CPU, I get ~10% frame rate increase in AC Odyssey benchmark and ~40% fps gain in the most taxing place in the game (middle of greater athens at rush hour). But far more importantly than that, I also get a much more constant frame time (in other words, no more stutter). In GR Wildlands I get a more modest 8% gain, but again, better frame time. As for older games, in Borderlands 2 I get ~35% more frames and once again, much better frame times. Bl2 has micro stutter even at 120fps, but with dxvk I have a butter smooth game play.

Dxvk is meant for Linux, but it also works on Windows. If you are going to try it, know that you need both d3d11.dll and dxgi.dll for dx11 games. Also note that you will see some stutter while the emulator builds its shader cache on first run (mostly for a couple of minutes of game play or during first benchmark run).
 
Got my 2700x for $129.99 at Micro Center,plus $30 off with new Asus x470 Prime MB, back in February, cost per frames ,not bad at all.
Got mine for €150 including tax, shipping, Borderlands 3 and the HSF. Hard to beat that.

Still, am quite surprised at the noticeable difference to the 3700x even with the 2060 super.

I disagree, it's faster if all you do is game, for testing esports it would have made more sense to load twitch in the background streaming at HD like alot of esports players these days.
I definitely agree on that. If I see all the apps my kid runs while gaming with the dual monitor setup, it is amazing how well the system handles this.

Seriously, would not want to give up this multi tasking capability and would love to see this benchmarked.

So playing a game while recording / streaming with discord, TeamSpeak and a browser running with YouTube, plus the various background tasks.
 
Starcraft 2 is optimized like a trash heap and it has frame drops no matter what CPU you use. It's not a good game to benchmark with.
Haha, yes but newer CPUs have higher clocks speeds which I thought would helps matters. I going to upgrade to Zen 3 once it comes out and I shall see then I guess.
 
What you are describing is 1% and 0.1% lows. Of which this review already contains 1% lows.
I don't have 1% lows displayed while gaming but I definitely get lower frames rates once the number of units increase (which is too be expected for a DX9 game).
 
4c4t CPUs (especially older ones) are going to suffer in some dx11-dx9 titles, old and new. Try dxvk (vulkan emulator). With a similar CPU, I get ~10% frame rate increase in AC Odyssey benchmark and ~40% fps gain in the most taxing place in the game (middle of greater athens at rush hour). But far more importantly than that, I also get a much more constant frame time (in other words, no more stutter). In GR Wildlands I get a more modest 8% gain, but again, better frame time. As for older games, in Borderlands 2 I get ~35% more frames and once again, much better frame times. Bl2 has micro stutter even at 120fps, but with dxvk I have a butter smooth game play.

Dxvk is meant for Linux, but it also works on Windows. If you are going to try it, know that you need both d3d11.dll and dxgi.dll for dx11 games. Also note that you will see some stutter while the emulator builds its shader cache on first run (mostly for a couple of minutes of game play or during first benchmark run).
Thanks, I'll try that out.
 
Based on just about every single benchmark test you have up there the choice is clear: Buy Intel.

The 10600k and 3700x are both just below $300 and within a $15 difference of each other.
For gaming, the 10600k performs similarly to the 9900k (the better overall CPU than both).
How much is the intel motherboard vs AMD? I can literally put that 3700X in my B350 motherboard that cost me £70.
 
I used to fight the FPS battle. Not any more. Too expensive. And AMD seems to bode well toward my styles. Middle ground GPU AMD holds it's own more often in titles I prefer, and even beats Intel often in those. I currently play on a 2070 GPU. I just don't see the extra dollars needed for Intel paying off for me. And, I'm pretty sure it won't pay off for most others either.
 
What you are describing is 1% and 0.1% lows. Of which this review already contains 1% lows.
If you think that then you don’t understand the CPUs role in a gaming PC. Gaming performance on a CPU is not consistent at all and varies quite widely. A 1% low might just represent a single moment within the game, it hides all the other times the frame rate can drop. If an average frame rate is 60 and the min is 30 how do you determine where between that your game would be? It could mean frequent drops to 45 with normal being 70, or could mean a consistent 60 with the fraction drop to 30, two very different experiences for the user. A graph showing frame times vs time can highlight the spikes, other tech reviewers have compiled this and the result is a line that looks like a mountain range. 1% lows doesn’t tell you anywhere near as much.
 
What is most disappointing about Intel 10th is their overclocking, they simply clock worse than 8th gen, what the heck ?
The majority of 8700K can reach 5.1ghz and above (HUB tested 10 retail samples and 7 of them reach 5.1ghz and above, Silicon lottery statistics also show ~70% of 8700K can reach 5.1ghz and above) while 10600K can barely reach 5.0ghz from all the reviews I saw and 5.1ghz show signs of instability in GamersNexus testing (lower 1% and 0.1%).

It's safe to assume that 8700K (or 8086K) is still the 2nd best gaming CPU beside the 10900K when you overclock them. And yes 8700K can clock higher than 9700K if anyone is asking.

Also only 20% of 10700K can reach 5.1ghz, that's just pathetic
SiliconLottery Statistics
 
Last edited:
If you think that then you don’t understand the CPUs role in a gaming PC. Gaming performance on a CPU is not consistent at all and varies quite widely. A 1% low might just represent a single moment within the game, it hides all the other times the frame rate can drop.

:facepalm: They are called 1% LOWS for a reason. That's plural. This is just embarrassing, so here's a video explaining.


1% and 0.1% lows take a sample across the entire testing period. They are representative of the frame dips throughout the entire session, not just one off drops.
 
Based on just about every single benchmark test you have up there the choice is clear: Buy Intel.

The 10600k and 3700x are both just below $300 and within a $15 difference of each other.
For gaming, the 10600k performs similarly to the 9900k (the better overall CPU than both).

Actually, there is about $100 difference between the two CPUs in my market (Atlanta, GA). Plus you get a much better platform on the AMD then you do the Intel. So not sure how you can say just buy Intel when the advantage is with AMD. As a LONG time PC user, I can tell you there is a LOT more to owning a computer then to gaming and higher FPS.
 
Last edited:
Back