On top of that, it kinda seems like this is a fix that is primarily targeting CPU bottlenecks, so mostly simulation-heavy games like CK3, the Civ series, Stellaris etc. This is relatively without consequence for games that are actively bottlenecked by graphics, and those are the games where the framerate improvements would be the most useful. But then again, it's an APU, maybe that just performs differently. I'm not really that well versed in how APUs work exactly.

Cool that it exists, don't get me wrong, but I don't think it's not going to have a lot of impact on the steamdeck. Would be cool to see more games tested tho.

[–] kogasa@programming.dev 7 points 1 day ago

Not CPU limited, actually the opposite. It targets games where the CPU is frequently waiting for the GPU to catch up. It fixes a defect in the AMD pstate driver's active (EPP guided) mode that would cause the CPU to be treated like it's idle if it's often taking short sleeps, even if it's actually quite busy.

[–] BakedCatboy@lemmy.ml 2 points 1 day ago

Vrchat is heavily CPU limited if you have a decent graphics card - me and my friends all have various x3d cpus and it's still not enough so this is welcome news. I have a 9800x3d and a friend has the 9850x3d and in tough situations we get around 50-70fps which is pretty middling for VR.

[–] thingsiplay@lemmy.ml 8 points 2 days ago

It's not just simulation heavy games that stress the CPU a lot. There are lot of single player games that would (or could) benefit from this. Things like enemy and character ai (not that LLM Ai of sort, I mean just the normal game logic) are stressing the CPU a lot and can cause stutter in certain regions of the games. It does not mean a constant stress, but like in Witcher 3, Baldur's Gate 3 or Crimson Desert in certain situations when its not about the graphics anymore (dungeon with lots of enemies, town with lots of people).

But we need more testing with those games before coming to a conclusion off course. We can just speculate at this point.

[–] LurkingLuddite@piefed.social 17 points 2 days ago* (last edited 2 days ago) (1 children)

Could boost 1% framerates by 30%+... way to fucking fail miserably at your ONE job...

[–] massive_bereavement@fedia.io 13 points 2 days ago (3 children)

What are you yapping about? The 1% refers to the low end of framerates, usually when games have lag spikes. This approach makes those spikes less noticeable because the APU power management doesn't get in the way.

[–] nekusoul@lemmy.nekusoul.de 26 points 2 days ago

If I had to guess this is about the blatantly false title of the article.

[–] Davel23@fedia.io 15 points 2 days ago (1 children)

I think he's berating the author of the article for using a title that implies that the improvement is to overall frame rates, not 1% lows.

[–] laurelraven@lemmy.zip 3 points 1 day ago

If I understand them correctly, doesn't boosting 1% lows have a bigger impact on the smoothness of the game than the overall? Boosting the high of 60 up to 80 if the lows are around 30 is still going to feel a bit jagged, but boosting that 30 to 40 is going to do a lot more to smooth it out.

I haven't read the article, being perfectly honest, but it targeting 1% lows makes this better, not worse, like OP seems to suggest