New Intel Linux Graphics Driver Patches Released, Up To 10-15% Better Performance

A new set of patches have been released for the Intel Linux graphics driver that "can provide 10-15% better performance when operating in the tuned mode," reports Phoronix. From the report: The set of Intel i915 Linux kernel graphics driver patches are about exposing the Intel RPS (Requested Power State) up/down thresholds. Right now the Intel Linux kernel driver has static values set for the up/down thresholds between power states while these patches would make them dynamically configurable by user-space. Google engineer Syed Faaiz Hussain raised the issue that they experimented with the Intel RPS tuning and were able to manage up to 15% better performance. With Counter-Strike: Global Offensive with OpenGL was a 14.5% boost, CS:GO with Vulkan was 12.9% faster, and Civilization VI with OpenGL was 11% faster while Strange Brigade was unchanged. No other game numbers were provided.

But as this is about changing the threshold for how aggressively the Intel graphics hardware switches power states, the proposed patches leave it up to user-space to adjust the thresholds as they wish. Google engineers are interested in hooking this into Feral's GameMode so that the values could be automatically tuned when launching games and then returning to their former state when done gaming, in order to maximize battery life / power efficiency. The only downside with these current patches are that they work only for non-GuC based platforms... So the latest Alder/Raptor Lake notebooks as well as Intel DG2/Alchemist discrete graphics currently aren't able to make use of this tuning option.

They need to be dynamic based on recent use, etc.

[dgatwood]

It seems to me that these sorts of limits should dynamically change, using the list of software that is running and recent past behavior to estimate the probability that a core or whatever will be needed at a particular power state within the next time quantum, estimating the badness of the impact if it is not available in time, and using levels in between to mitigate the badness of the transition.

More to the point, frequent "frame not ready" failures should provide a feedback loop that automatically lower

questions

[danda]

1. Would this speed up video playback or at least make it more efficient? My old thinkpad x220 gets hot when playing a long video, or some hi-res vids even stutter/pause.

2. Would these driver patches work on an old distro, eg ubuntu 16.04, 18, etc? Or have to be running bleeding edge kernel?

thx.

i915 is a pretty old driver codebase. Good.

[Seven Spirals]

I'm glad to see this. I'm not sure how it'll impact older systems, but those are the ones I care most about in this situation. There are so many low-end systems with i915-derived graphics. This is great! There was an old change to the driver that resulted in huge speed gains on the FB console a few years ago. I loved it. I spend a lot of time there and it was very welcome. It also made it to FreeBSD!

15% a big deal?

[Latent Heat]

Is Moore's Law giving a doubling in processor power every 18 months dead, so now a 15% performance boost is meaningful?

Or is it "15% here, 15% there and pretty soon you have a serious performance boost?"

Re:15% a big deal?

[Vlad_the_Inhaler]

At the risk of stating the obvious, Moore's Law applies to hardware, this is a software update which speeds existing hardware up.