Linux Gaming Performance With AMD Ryzen 5 2600X / Ryzen 7 2700X

[phoronix]

Phoronix: Linux Gaming Performance With AMD Ryzen 5 2600X / Ryzen 7 2700X

Today the Ryzen+ "Pinnacle Ridge" processors begin shipping and we can now share with you the initial performance results for the Ryzen 5 2600X and Ryzen 7 2700X processors. One of the most common questions I've received about these improved Zen processors since showing them off last week was inquiries/hopes about the Linux gaming performance, so those numbers are first up today followed by other Linux benchmark results forthcoming.

http://www.phoronix.com/vr.php?view=26223

[Dukenukemx]

I was hoping for more performance but it seems to be within the margin of error compared to the 1800X.

[schmidtbag]

Strange how gaming performance still isn't all that great.

Thanks for posting the $/performance graphs - not enough reviewers do this.

[cybertraveler]

I wonder if Serious Sam 3 is heavily and efficiently multi-threaded, thus giving the Ryzen's a better chance. I guess the Intel chip wont go into the boosted high frequency mode if many of its cores are being worked hard.

[pinguinpc]

Michael

Good article and confirm intel is best 1080p (or lower) game cpu

[monte84]

Michael I know with my 1700 when using the performance governor and not manually overclocking, my CPU would be forced to the base clock speeds and not the turbo speed, I would assume this is a similar situation with the 2700X. If MCE (Multi Core Enhancement) is enabled on the Intel system (default on most aftermarket board) then this could be a 1GHz difference in clock speed vs the 2700X and the 8700K which goes a long way into explaining such a large gap and not a good representation of performance for either. Can you confirm either of these or try clocking the CPU to similar speeds?

[humbug]

Strange how gaming performance still isn't all that great.

It seems to be a software issue on Linux.

Look at the windows numbers

https://www.pcper.com/reviews/Processors/Ryzen-7-2700X-and-Ryzen-5-2600X-Review-Zen-Matures/Gaming-Performance

https://www.pcper.com/reviews/Processors/Ryzen-7-2700X-and-Ryzen-5-2600X-Review-Zen-Matures/1440p-and-4K-Gaming

Do the porting frameworks used make the Linux games less well multithreaded? Which swings the pendulum towards Intel?

[monte84]

It seems to be a software issue on Linux.

Look at the windows numbers
https://www.pcper.com/reviews/Proces...ng-Performance
https://www.pcper.com/reviews/Proces...-and-4K-Gaming

Do the porting frameworks used make the Linux games less well multithreaded? Which swings the pendulum towards Intel?

I think it is because we are seeing a close to 1GHz difference in clock speed, as I stated above. With intel MCE it allows the turbo speed to all cores which is 4.7Ghz, assuming cooling and power delivery requirements are met. When using performance governor on Ryzen the clock speed is forced to the base which is 3.7Ghz. This would also mean there is only a 100Mhz difference between the 1800X and the 2700X, as neither are hitting turbo speeds.

[sa666666]

Michael I know with my 1700 when using the performance governor and not manually overclocking, my CPU would be forced to the base clock speeds and not the turbo speed, I would assume this is a similar situation with the 2700X. If MCE (Multi Core Enhancement) is enabled on the Intel system (default on most aftermarket board) then this could be a 1GHz difference in clock speed vs the 2700X and the 8700K which goes a long way into explaining such a large gap and not a good representation of performance for either. Can you confirm either of these or try clocking the CPU to similar speeds?

If this is indeed the case, how should a Ryzen system be configured to get the best performance? Right now I always use 'Performance', but I'd like it to be optimal, of course.

[carewolf]

Why compare against then Intel+AMD hybrid instead of a real Intel CPU?