For Honor: Graphics & CPU Performance

Thanks for another in-depth performance analysis, Steve. I'm kind of surprised to see how well the i5 2500k performed.
 
Great article. Thanks Techspot!
This game looks very nice, I might buy it for a little hack and slash fun.

Game seems to scale pretty well with different CPUs. Shocking to see how the i3s are beating AMD , again...
We only have to wait a little more for Ryzen. Hopefully the landscape will change.

Cheers
 
Surprised to see the gtx 970 beaten by the 1060 6gb by a considerable margin! Thought they performed the same but obviously not in the benchmarks.
 
The i5-7600K outperforms the i7-6700K? That's ... unexpected. Based on clock speed, cores, cache, IPC... it shouldn't.

Maybe For Honor uses HEVC 10bit decode instructions somehow???
 
The i5-7600K outperforms the i7-6700K? That's ... unexpected. Based on clock speed, cores, cache, IPC... it shouldn't.
Seems to be using an older game engine/no more then 4 logical cores, common for games using older tech; its not even close to taking advantage of what CPU's have to offer. Look at how good the Core i3-6100 is doing, its keeping pace with an i7-5960X.
Not a good game to compare CPU performance, Gears 4 and Overwatch are better suited for that. As far as the game itself, looks like a copy of Chivalry: Medieval Warfare.
 
Surprised to see the gtx 970 beaten by the 1060 6gb by a considerable margin! Thought they performed the same but obviously not in the benchmarks.

The 6 GB version of 1060 should really be called 1060 Ti when compared to its 3 GB version.
 
Why did you choose not to use the in game benchmark for the CPU test?

Seems like your adding a large margin for error without doing so?
 
Just bought the 1070 from EVGA. Now I have another excuse to upgrade the AMD FX-8320. Hopefully the Ryzen will deliver. If not, guess it's time for Intel.
 
Why did you choose not to use the in game benchmark for the CPU test?

Seems like your adding a large margin for error without doing so?

Does this cover it?

"I did find that the processor was taxed much more heavily when actually playing the game and that being the case the CPU numbers are based on actual game play during what I found to be a particularly CPU-demanding scene."
 
Why did you choose not to use the in game benchmark for the CPU test?

Seems like your adding a large margin for error without doing so?

Does this cover it?

"I did find that the processor was taxed much more heavily when actually playing the game and that being the case the CPU numbers are based on actual game play during what I found to be a particularly CPU-demanding scene."

Thanks for the reply. I guess I always read the CPU charts as a i5 vs i7 vs. the rest shootout, but your actually providing an example of what a player can expect to see realtime with their respective hardware. I get it, I still think manually playing through scenes is less of an apples to apples comparison than a pre determined bench.
 
Thanks for the reply. I guess I always read the CPU charts as a i5 vs i7 vs. the rest shootout, but your actually providing an example of what a player can expect to see realtime with their respective hardware. I get it, I still think manually playing through scenes is less of an apples to apples comparison than a pre determined bench.

The vast majority prefer us to report results based on a custom pass using Fraps for example. Canned benchmarks can be misleading. In this case using the canned benchmark for testing CPUs is utterly useless.
 
Nice benchmark review. I recently upgraded to an RX 480 and glad to know its performing in league with the GTX 1060. I wish the new games come with DX 12 and Vulkan support this year.

Pathetic CPU performance by aging FX series. Lets hope the Ryzen performs and costs well.
 
Why is the minimum frame rate taking such a hit from your CPU overclocking tests, compared to the non overclocking results? The 5ghz 7700k overclock shows to be giving worse minimum frames than the 2500k at stock speed of 3.3 GHz! The only benefit is 2 extra fps at the top end, which is not at all worth the massive hit to minimum fps. You say in the article that you recommend overclocking to get the most out of the game, but from the results you post I would say the opposite is true! Are you running different tests for the overclocking and stock tests?
 
Why is the minimum frame rate taking such a hit from your CPU overclocking tests, compared to the non overclocking results? The 5ghz 7700k overclock shows to be giving worse minimum frames than the 2500k at stock speed of 3.3 GHz! The only benefit is 2 extra fps at the top end, which is not at all worth the massive hit to minimum fps. You say in the article that you recommend overclocking to get the most out of the game, but from the results you post I would say the opposite is true! Are you running different tests for the overclocking and stock tests?

It now looks like the benchmark in the article has been updated with new minimums showing for the CPU overclock benchmark, but no note about this edit in the article.
 
It now looks like the benchmark in the article has been updated with new minimums showing for the CPU overclock benchmark, but no note about this edit in the article.

It was fixed long ago, possibly a cache issue with either your PC or the server.
 
Great article. Thanks Techspot!
This game looks very nice, I might buy it for a little hack and slash fun.

Game seems to scale pretty well with different CPUs. Shocking to see how the i3s are beating AMD , again...
We only have to wait a little more for Ryzen. Hopefully the landscape will change.

Cheers
With this game at least it likely wont. Seeing as 2500k and 5960x that both have very similar single thread performance are performing near identically but one being a 4c4t CPU and another being 8c16t, it is easy to conclude that this game is not made to properly utilize anything past 4 threads. It being a reasonable expectation that 7700k will hold the title as the processor with the highest single thread performance and being a 4c8t CPU, I have hard time imagining any 4/6/8 core zens surpassing its performance, in this game at least. Since extreme settings are already keeping titan xp overworked and 7700k offers diminishing gains above 3.5GHz zen should be able to perform on par with 7700k in this particular scenario, but say you wanted to play at 144fps for high refresh rate monitors, you would have to lower your graphical settings quite a bit to alleviate the GPU bottleneck and your CPU would again be more relevant.
 
Back