Tuesday, December 6th 2016

AMD Cripples Older GCN GPUs of Async-Compute Support?

AMD allegedly disabled asynchronous-compute technology support on older generations of Graphics CoreNext (GCN) architecture, since Radeon Software 16.9.2. With the newer drivers, "Ashes of the Singularity" no longer supports asynchronous-compute, a feature that improves performance in the game, on GPUs based on the first-generation GCN architecture, such as the Radeon R9 280X.

"Ashes of the Singularity" benchmarks run by Beyond3D forum members on GCN 1.0 hardware, comparing older drivers to version 16.9.2 shows that the game supports async-compute on the older drivers, and returns improved performance. AMD, on its part, is pointing users to a patch change-list from the developers of "Ashes..." which reads that the game supports DirectX 12 async-compute only on GCN 1.1 (eg: Radeon R9 290) and above.
Source: Reddit
Add your own comment

97 Comments on AMD Cripples Older GCN GPUs of Async-Compute Support?

#51
jigar2speed
Tarun again ?
renz496you know what nvidia did support their gpu longer than AMD. right now those fermi still get driver updates from main driver. AMD cards from the same generation already being push into legacy status by AMD.
Yeah Fermi has full DX 12 support... Oh wait...
Posted on Reply
#52
renz496
jigar2speedYeah Fermi has full DX 12 support... Oh wait...
oh please. don't try change the subject. this is not about DX12. i'm talking about pure driver support on older product. AMD droping driver support much early than nvidia is pure fact. remember when AMD dropping support for 4k series after 3 year plus? if AMD still rapidly pumping new architecture like they did before GCN 1.0 probably already on legacy status.

that aside i see some people try to put this claim to test at beyond 3d.
Posted on Reply
#53
HD64G
bugDo you know that for a fact? Are you 100% sure they couldn't keep the old codebase around and extract extra performance from the newer GPUs?

As in, you know:
if(GCN>=1.1) {
doNewMagic();
} else {
doWhatYouWereDoingAlready();
}

Edit: also, where's the post that claims support was dropped, it's not linked in the article...
Saying dropping support means "stop increasing performance through drivers for CGN gen1" in my view.
Posted on Reply
#54
ShurikN
Looking at this article and the clickbait title as well as that HDR over HDMI one a 10 days ago, no wonder there are such a small amount of AMD cards sent for review compared to nV.
Stay classy TPU.
Posted on Reply
#55
the54thvoid
Super Intoxicated Moderator
Can anyone confirm with evidence the beyond3d results? If a driver roll back on GCN 1 cards delivers higher fps, it nullifies claims it was a patch or infers AMD acquiesced to a removal of support.
Or the patch was requested by AMD? A larger selection of games is required to test the notion of driver based decreases.
Either way, it's a non issue, given the age of cards and small game sample.
Posted on Reply
#56
Ubersonic
ADHDGAMINGCalm down this was the Devs move not AMDs ... Read the article not just the headline .
O.o

AMD released drivers, installing those drivers disables AC on older cards in AotS, rolling back to the older drivers re-enables AC on older cards in AotS. How exactly is this supposed to be a dev issue? did they write the AMD drivers?

Granted supporting a half decade old card in a cutting age game may be undesirable but this is still a bum move by AMD.
Posted on Reply
#57
Aquinus
Resident Wat-man
EarthDogConfused... titles asks a question... yet the body of the news says it wasn't AMD its the Ashes devs...........?
Sounds a lot like this thread: www.techpowerup.com/forums/threads/amd-radeon-gpus-limit-hdr-color-depth-to-8bpc-over-hdmi-2-0.227941/

Where the title says AMD gimped their cards when using HDMI 2.0 and HDR when HDMI 2.0 is the problem, would exist on anything using HDMI 2.0, and just sucks compared to DisplayPort... but @btarunr thought pointing a finger at AMD with the title was a good idea. Now I see this thread, once again, pointing a finger at AMD when it very well could be the application at fault for making technical design decisions. Simply put, what the hell, TPU?!
Ubersonico_O

AMD released drivers, installing those drivers disables AC on older cards in AotS, rolling back to the older drivers re-enables AC on older cards in AotS. How exactly is this supposed to be a dev issue? did they write the AMD drivers?

Granted supporting a half decade old card in a cutting age game may be undesirable but this is still a bum move by AMD.
...because games can change what they're doing based on a driver's version or the kind of hardware present. It's not uncommon for devs to do this to take advantage of features on different kinds of GPUs but, still having something to fall back on when the hardware doesn't support it or if it's not advantageous to use it.
Posted on Reply
#58
AlienIsGOD
Vanguard Beta Tester
i thought news editors were supposed to be impartial? these Anti AMD headlines do nothing other than serve one mans opinion.
Posted on Reply
#59
TheinsanegamerN
jigar2speedTarun again ?


Yeah Fermi has full DX 12 support... Oh wait...
And the 5000 series hasnt gotten a full driver since 15.7 (or that beta 16.2 if you like things breaking). Neither has the 6000 series. Meanwhile that 400 or 500 series card is on the latest nvidia driver.

So would you rather have no DX12, or not be able to play anything released in DX11 in the last 2 years, PLUS no DX12?
Posted on Reply
#60
FordGT90Concept
"I go fast!1!11!1!"
HD 5000 and HD 6000 are pre-GCN cards. They are on legacy product support.
Posted on Reply
#61
Kwee
I see a lot of misunderstand since i post my finding.

In a first place, i started to search why GCN 1.0 was not supported in Rise Of The Tomb Raider last patch.

When Maxwell was accused to not support Async Compute, someone create a program for testing Async Compute. It worked on GCN 1.0, 1.1 and 1.2 back in time.

So i just wanted to verify if that still the case and that how i find that Async Compute was disabled on news drivers. Then i post on Reddit.

After that, many ask me to test a game that support Async Compute on GCN 1.0. So i tried Ashes Of Singularity. You know the end. That just confirm that Async Compute was disabled on news drivers. Old drivers performs way better because of Async Compute.

DirectX12 driver 16.3.1 Async Compute off : i.imgur.com/aiV1pSg.png

DirectX12 driver 16.3.1 Async Compute on : i.imgur.com/CGrb4yM.png

DirectX12 drivers superior to 16.9.2 Async Compute off :i.imgur.com/yiSSRCE.png

DirectX12 drivers superior to 16.9.2 Async Compute on :i.imgur.com/Fch5V8w.png
Posted on Reply
#62
TheinsanegamerN
FordGT90ConceptHD 5000 and HD 6000 are pre-GCN cards. They are on legacy product support.
That doesn't stop people from saying "but but but NVIDIA DRIVER SUPPORT IS TERRIBLE" and ignoring that AMD outright dropped their older cards, while nvidia still supports the 400 and 500 series.

Nobody is arguing that the 5000 and 6000 series isnt legacy.
Posted on Reply
#63
bug
KweeI see a lot of misunderstand since i post my finding.

In a first place, i started to search why GCN 1.0 was not supported in Rise Of The Tomb Raider last patch.

When Maxwell was accused to not support Async Compute, someone create a program for testing Async Compute. It worked on GCN 1.0, 1.1 and 1.2 back in time.

So i just wanted to verify if that still the case and that how i find that Async Compute was disabled on news drivers. Then i post on Reddit.

After that, many ask me to test a game that support Async Compute on GCN 1.0. So i tried Ashes Of Singularity. You know the end. That just confirm that Async Compute was disabled on news drivers. Old drivers performs way better because of Async Compute.

DirectX12 driver 16.3.1 Async Compute off : i.imgur.com/aiV1pSg.png

DirectX12 driver 16.3.1 Async Compute on : i.imgur.com/CGrb4yM.png

DirectX12 drivers superior to 16.9.2 Async Compute off :i.imgur.com/yiSSRCE.png

DirectX12 drivers superior to 16.9.2 Async Compute on :i.imgur.com/Fch5V8w.png
Thanks, I think this sheds a lot more light on the issue.
Posted on Reply
#64
wiak
stop blaming amd for what game developers are obviously doing, amd has less control over dx12/vulkan than they did with dx9/dx10/dx11/opengl
Posted on Reply
#65
Kwee
wiakstop blaming amd for what game developers are obviously doing, amd has less control over dx12/vulkan than they did with dx9/dx10/dx11/opengl
It's a driver issue, not a game issue.
Posted on Reply
#66
jigar2speed
renz496oh please. don't try change the subject. this is not about DX12. i'm talking about pure driver support on older product. AMD droping driver support much early than nvidia is pure fact. remember when AMD dropping support for 4k series after 3 year plus? if AMD still rapidly pumping new architecture like they did before GCN 1.0 probably already on legacy status.

that aside i see some people try to put this claim to test at beyond 3d.
Dude, i am not sure what you are smoking but i really like that stuff.
Posted on Reply
#67
R-T-B
jigar2speedDude, i am not sure what you are smoking but i really like that stuff.
What part of what he said was untrue? I did not catch anything.
Posted on Reply
#68
bug
R-T-BWhat part of what he said was untrue? I did not catch anything.
The part jigar2speed doesn't like, obviously.
Posted on Reply
#69
ADHDGAMING
AquinusSounds a lot like this thread: www.techpowerup.com/forums/threads/amd-radeon-gpus-limit-hdr-color-depth-to-8bpc-over-hdmi-2-0.227941/

Where the title says AMD gimped their cards when using HDMI 2.0 and HDR when HDMI 2.0 is the problem, would exist on anything using HDMI 2.0, and just sucks compared to DisplayPort... but @btarunr thought pointing a finger at AMD with the title was a good idea. Now I see this thread, once again, pointing a finger at AMD when it very well could be the application at fault for making technical design decisions. Simply put, what the hell, TPU?!

...because games can change what they're doing based on a driver's version or the kind of hardware present. It's not uncommon for devs to do this to take advantage of features on different kinds of GPUs but, still having something to fall back on when the hardware doesn't support it or if it's not advantageous to use it.
THANK YOU Aquinus
Posted on Reply
#70
AsRock
TPU addict
Sempron Guywell considering anything lower than the 280x is GCN 1.0 which is said particular "devs" are excluding for async support, maybe it's time for an upgrade? The 280x is what 3 years old now?
And when was the 290X released ? :P ;), although wasn't the 280 a tweaked 7970 or some thing like that, so yeah it's older than that but it's still GCN and devs should support it for a few more years at least as the card is not a bad card still today.

I think this is less AMD and more lazy devs.
Posted on Reply
#71
rtwjunkie
PC Gaming Enthusiast
AsRockalthough wasn't the 280 a tweaked 7970
Close. IIRC, and I may not, the 280x was the tweaked 7970, that IMHO was not as good as a 7970.
Posted on Reply
#72
Tatty_Two
Gone Fishing
ADHDGAMINGTHANK YOU Aquinus
Apart from the fact that since Aquinus's post the author of the reddit comments has come and explained his findings, after his tests he does believe it's a driver issue.
Posted on Reply
#73
Aquinus
Resident Wat-man
Tatty_OneApart from the fact that since Aquinus's post the author of the reddit comments has come and explained his findings, after his tests he does believe it's a driver issue.
It's an indicator but, it's certainly not definitive as you can check to make sure that a version is under a certain version. For example, he said he tested:
KweeDirectX12 driver 16.3.1 Async Compute off : i.imgur.com/aiV1pSg.png

DirectX12 driver 16.3.1 Async Compute on : i.imgur.com/CGrb4yM.png

DirectX12 drivers superior to 16.9.2 Async Compute off :i.imgur.com/yiSSRCE.png

DirectX12 drivers superior to 16.9.2 Async Compute on :i.imgur.com/Fch5V8w.png
16.3.1 to 16.9.2 is a really big gap and if an update on a minor version was when it stopped working, it could still be the case that games looking to make sure that certain features are used when the driver is in a certain range. I don't want to make assumptions about software though. Simple fact is that we don't really know. An interesting observation though is that the 16.9.2 results are somewhere between the 16.3.1 results.

Either way, I think more testing is in order to determine if AMD actually gimped their drivers or not or if AMD gimped async compute like how they gimped HDMI. ;)
Posted on Reply
#74
Kwee
I spent all night installing every drivers between 16.3.1 and 16.9.2. The break point is the driver 16.4.2(released in April). After this driver no more Async Compute on GCN 1.0. So Nixxes was aware that Async Compute was not active on GCN 1.0. They released Async Compute patch for Rise Of The Tomb Raider in July, specifying that only GCN 1.1 and superior can take advantage of Async Compute.

I received a good amount of results all around the world that users did for me and their results just confirmed my findings. Thank you all.
Posted on Reply
#75
Sempron Guy
AsRockAnd when was the 290X released ? :p ;), although wasn't the 280 a tweaked 7970 or some thing like that, so yeah it's older than that but it's still GCN and devs should support it for a few more years at least as the card is not a bad card still today.

I think this is less AMD and more lazy devs.
well as you said 280x is nothing more than 7970 ghz ed. 290x is a newer architecture. 4 years lifespan of a card is more than enough. AMD is not running a charity nor do they have the luxury to do so. That effort to still support a 4 years old architecture is best spent elsewhere.
Posted on Reply
Add your own comment
Nov 21st, 2024 08:58 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts