RADV Gets A Performance Improving Workaround For Vega, Helps Dota 2

[boxie]

We did not. Vega has a hardware bug with 2 known workarounds:

1) simple but can have a performance cost.
2) less simple but almost no performance cost.

We previously thought the cost by workaround 1 was negligible and hence it was not worth implementing 2, but I noticed that rendering the GUI in dota2 there was worst case behavior for workaround 1 which caused some slowdown.

Part of the difficulty finding this was that my CPU is worse than what Michael uses for his benchmarks, so I'm CPU limited on 4k with my Vega, and hence did not see the performance difference until I forced the GPU to low power.

Hey man, just wanted to say - Thanks for all your great work - you are a legend, a gentleman and a scholar!

~ A very happy vega owner

[abott]

I got the impression because as i understood it , the article says that dota is cpu limited because of using the api irresponsibly , or at least not optimally , which happened to impact Vega's performance particularly bad ; apparently that is not the case ( should source have source 2 have been better optimized anyways ? ) but either way this is an ad hoc "solution"

You are being an idiot. This is a hardware bug, listed clearly in the commit, and the solution affects nothing else as they test for a hardware flag that is added only to affected hardware. This is not adhoc. Actually, this is code you'd never see in production because it's very clean and well written, along with properly documented and clear in the documentation.

You shouldn't be commenting this garbage on stuff when you can't understand something this basic. This is a good improvement that polishes the entire stack by making them all perform to their best, and nothing else.

[bridgman]

I got the impression because as i understood it , the article says that dota is cpu limited because of using the api irresponsibly , or at least not optimally , which happened to impact Vega's performance particularly bad ; apparently that is not the case ( should source have source 2 have been better optimized anyways ? ) but either way this is an ad hoc "solution"

You are being an idiot. This is a hardware bug, listed clearly in the commit, and the solution affects nothing else as they test for a hardware flag that is added only to affected hardware. This is not adhoc. Actually, this is code you'd never see in production because it's very clean and well written, along with properly documented and clear in the documentation.

You are both right, except for the part about GunpowaderGuy being an idiot. It was a hardware bug, but the simpler workaround was considered very low overhead until Baz noticed that that DOTA 2 was calling into the affected API function much more than other apps, to the point where the simpler workaround was contributing some noticeable overhead.

It's not clear at this point whether the API function usage by DOTA2 would be considered excessive or not.

[pal666]

You are both right, except for the part about GunpowaderGuy being an idiot.

if last part is not clear from this thread, just look at systemd threads

[abott]

Didn't call him an idiot, said he was being an idiot. He can at least fix that one. If he couldn't, I would just call him an idiot.

[TemplarGR]

You are being an idiot. This is a hardware bug, listed clearly in the commit, and the solution affects nothing else as they test for a hardware flag that is added only to affected hardware. This is not adhoc. Actually, this is code you'd never see in production because it's very clean and well written, along with properly documented and clear in the documentation.

You shouldn't be commenting this garbage on stuff when you can't understand something this basic. This is a good improvement that polishes the entire stack by making them all perform to their best, and nothing else.

Why this forum is full of people like this who are blatantly toxic to others? Have some decency people...

[Guest]

One of the points of vulkan was ending this ad hoc hackery , if the inefiency is the game's fault then the driver shouldn't bend over backwards to fix it , just expose a good way ( performance and closeness to metal /how much it diffcults cross hardware support ) to deal with it

But... business. Gamers wanna play games, they don't care about drivers.

[abott]

Why this forum is full of people like this who are blatantly toxic to others? Have some decency people...

That was me being decent. Spew garbage, get garbage.

[VikingGe]

That was me being decent.

If you actually believe that, I really hope that you have no contact to any life form in the real world. You're not being "decent", you're being the incarnation of everything that's wrong about humanity these days.

What exactly makes you think that "you're an idiot" is a better way to start a sentence rather than simply saying "you're wrong"?

[abott]

If you actually believe that, I really hope that you have no contact to any life form in the real world. You're not being "decent", you're being the incarnation of everything that's wrong about humanity these days.

What exactly makes you think that "you're an idiot" is a better way to start a sentence rather than simply saying "you're wrong"?

Quote me correctly, "you're being an idiot." just like you are currently. Spew garbage, get garbage.