Jump to content
You must now use your email address to sign in [click for more info] ×

AMD Radeon RX Hardware Acceleration


Recommended Posts

It's seem working fine on latest driver? Has AMD fixed the issue? I recently purchased affinity designer on pc, just tried the 1.9.0.932(which still allow amd rx5700 opencl acceleration), the gpu acceleration seems working fine to me, much faster than without gpu acceleration enable. 

The problem tho is I cant install both 1.9.0.932 and latest affinity designer in the same pc, is there way to fix this?

Link to comment
Share on other sites

Hi, @dnwls

The short version: No.

 

AMD Ryzen 7 5700X | INTEL Arc A770 LE 16 GB  | 32 GB DDR4 3200MHz | Windows 11 Pro 23H2 (22631.3296)
AMD A10-9600P | dGPU R7 M340 (2 GB)  | 8 GB DDR4 2133 MHz | Windows 10 Home 22H2 (1945.3803) 

Affinity Suite V 2.4 & Beta 2.(latest)
Better translations with: https://www.deepl.com/translator  
Interested in a robust (selfhosted) PDF Solution? Have a look at Stirling PDF

Life is too short to have meaningless discussions!

Link to comment
Share on other sites

7 hours ago, dnwls said:

It's seem working fine on latest driver? Has AMD fixed the issue? I recently purchased affinity designer on pc, just tried the 1.9.0.932(which still allow amd rx5700 opencl acceleration), the gpu acceleration seems working fine to me, much faster than without gpu acceleration enable. 

The problem tho is I cant install both 1.9.0.932 and latest affinity designer in the same pc, is there way to fix this?

You could use one beta version in parallel to one retail version. But Affinity recommends against using beta versions for productive use, as they often contain bugs fixed in later versions, and file format could be changed, so later retail version refuse to offer some beta documents.

There are some betas almost identical to retail versions, but using them is on your own risk.

while HW acceleration is faster in many cases, some filters or e.g. blend ranges lead to Photo falls back to unaccelerated. It really depends on your workflow.

Mac mini M1 A2348 | Windows 10 - AMD Ryzen 9 5900x - 32 GB RAM - Nvidia GTX 1080

LG34WK950U-W, calibrated to DCI-P3 with LG Calibration Studio / Spider 5

iPad Air Gen 5 (2022) A2589

Special interest into procedural texture filter, edit alpha channel, RGB/16 and RGB/32 color formats, stacking, finding root causes for misbehaving files, finding creative solutions for unsolvable tasks, finding bugs in Apps.

 

Link to comment
Share on other sites

Well I have latest Affinity Photo and have just updated the AMD driver (end of June) on my RX6600 - no difference. STILL won't work if I try to tick hardware acceleration. Still waiting for action / feedback from somebody................AMD or Affinity.

Link to comment
Share on other sites

Just now, Mel_P said:

Well I have latest Affinity Photo and have just updated the AMD driver (end of June) on my RX6600 - no difference. STILL won't work if I try to tick hardware acceleration. Still waiting for action / feedback from somebody................AMD or Affinity.

The latest Affinity release (1.10.5) still has the restriction, so once AMD has a working driver you will need an updated Affinity release to support it. And once it is supported, I'm sure Serif will let us know.

-- Walt
Designer, Photo, and Publisher V1 and V2 at latest retail and beta releases
PC:
    Desktop:  Windows 11 Pro, version 23H2, 64GB memory, AMD Ryzen 9 5900 12-Core @ 3.00 GHz, NVIDIA GeForce RTX 3090 

    Laptop:  Windows 11 Pro, version 23H2, 32GB memory, Intel Core i7-10750H @ 2.60GHz, Intel UHD Graphics Comet Lake GT2 and NVIDIA GeForce RTX 3070 Laptop GPU.
iPad:  iPad Pro M1, 12.9": iPadOS 17.4, Apple Pencil 2, Magic Keyboard 
Mac:  2023 M2 MacBook Air 15", 16GB memory, macOS Sonoma 14.4

Link to comment
Share on other sites

I have been evaluating the compile latency of AMD drivers, using the code posted by the developers at github for a few months now (since i noticed the issue myself).

https://github.com/MarkIngramUK/ocl-compile-benchmark

My experimentation in Windows 11, starting from driver Adrenalin 21.10.2 all the way up the most recent Pro 22Q2 shows very reasonable latency with the latest (22Q2) being around 62-63ms on an RX6900. The best result on Windows was with the default Win11 driver (no AMD driver installed) with a latency slightly above 53ms, while with a small modification (of time measurement code) the same code can run in Linux. I tested it with the open source driver (in linux-kernel 5.17-5.18) the latency I measured was around 22ms. So, no matter how hard i tried to reproduce the issue (out of my own academic curiosity); i couldn't get any latency in the order of magnitude mentioned (of around 1400ms) in the github post (for the Radeon 5000 series).

Sadly it also seems that all recent Affinity releases have this restriction, so there is now way to enable HW acceleration in the application and test it, if they don't remove this restriction in a future release. We can only hope that they address this issue at least in their next major release (v.2). And i personally don't see a reason for it any longer, although i understand that there might have been an issue affecting several users at a given point in time using a specific system configuration (OS, driver, etc). Even if for specific configurations this might still be the case, i believe they should make it at least possible to enable the hardware acceleration (even if we need to acknowledge the risk and modify some configuration file) in any OpenCL capable device, so that we can also measure the performance from the benchmark as well as in real test cases in the application and provide feedback back to the developers.

Link to comment
Share on other sites

3 hours ago, kkoukos said:

I have been evaluating the compile latency of AMD drivers, using the code posted by the developers at github for a few months now (since i noticed the issue myself).

https://github.com/MarkIngramUK/ocl-compile-benchmark

My experimentation in Windows 11, starting from driver Adrenalin 21.10.2 all the way up the most recent Pro 22Q2 shows very reasonable latency with the latest (22Q2) being around 62-63ms on an RX6900. The best result on Windows was with the default Win11 driver (no AMD driver installed) with a latency slightly above 53ms, while with a small modification (of time measurement code) the same code can run in Linux. I tested it with the open source driver (in linux-kernel 5.17-5.18) the latency I measured was around 22ms. So, no matter how hard i tried to reproduce the issue (out of my own academic curiosity); i couldn't get any latency in the order of magnitude mentioned (of around 1400ms) in the github post (for the Radeon 5000 series).

Sadly it also seems that all recent Affinity releases have this restriction, so there is now way to enable HW acceleration in the application and test it, if they don't remove this restriction in a future release. We can only hope that they address this issue at least in their next major release (v.2). And i personally don't see a reason for it any longer, although i understand that there might have been an issue affecting several users at a given point in time using a specific system configuration (OS, driver, etc). Even if for specific configurations this might still be the case, i believe they should make it at least possible to enable the hardware acceleration (even if we need to acknowledge the risk and modify some configuration file) in any OpenCL capable device, so that we can also measure the performance from the benchmark as well as in real test cases in the application and provide feedback back to the developers.

Agreed. I've said this several times that the option should be put back for those who want to try anyway, with it defaulting to off.

Link to comment
Share on other sites

On 7/3/2022 at 5:35 AM, dnwls said:

It's seem working fine on latest driver? Has AMD fixed the issue? I recently purchased affinity designer on pc, just tried the 1.9.0.932(which still allow amd rx5700 opencl acceleration), the gpu acceleration seems working fine to me, much faster than without gpu acceleration enable. 

The problem tho is I cant install both 1.9.0.932 and latest affinity designer in the same pc, is there way to fix this?

wait what? it works with older affinity versions?

Current Workstation:
CPU: AMD Ryzen 5 5500 - MOBO: Asus B450 - RAM: 16GB DDR4 2667Mhz - GPU: AMD Radeon 7850 1GB
NVMe SSD: Crusial P3 1TB M.2 -  SSD: Samsung Evo 850 256GB  - PSU: XFX TS450 - OS: Win10

Link to comment
Share on other sites

@nitro912gr yes it works with the older version(v1.9.0.932), i believe rx5000 support removed after version 1.9.1(?).

I'm not sure if everything work, so far I only tried some simple drawing with brush tool, opencl acceleration seems working fine to me. The raster brush was significantly faster, barely any delay/lag when opencl enabled. 

Link to comment
Share on other sites

1 hour ago, nitro912gr said:

wait what? it works with older affinity versions?


introduced in 1.9.0 on 2021-02-04

removed for affected AMD gpu in 1.9.1 on 2021-02-25

 

Mac mini M1 A2348 | Windows 10 - AMD Ryzen 9 5900x - 32 GB RAM - Nvidia GTX 1080

LG34WK950U-W, calibrated to DCI-P3 with LG Calibration Studio / Spider 5

iPad Air Gen 5 (2022) A2589

Special interest into procedural texture filter, edit alpha channel, RGB/16 and RGB/32 color formats, stacking, finding root causes for misbehaving files, finding creative solutions for unsolvable tasks, finding bugs in Apps.

 

Link to comment
Share on other sites

oh yeah sorry, this problem is going for so long that I totally forgot that they disabled it due to poor performance...

But I agree we should get the option back and see for ourselves if it worth it. I believe in my scenario at home that I have a Ryzen 2600 and Radeon 5500XT I will get better performance no matter what.

Current Workstation:
CPU: AMD Ryzen 5 5500 - MOBO: Asus B450 - RAM: 16GB DDR4 2667Mhz - GPU: AMD Radeon 7850 1GB
NVMe SSD: Crusial P3 1TB M.2 -  SSD: Samsung Evo 850 256GB  - PSU: XFX TS450 - OS: Win10

Link to comment
Share on other sites

1 hour ago, nitro912gr said:

oh yeah sorry, this problem is going for so long that I totally forgot that they disabled it due to poor performance...

But I agree we should get the option back and see for ourselves if it worth it. I believe in my scenario at home that I have a Ryzen 2600 and Radeon 5500XT I will get better performance no matter what.

Agreed. We should get an official response as to why we can't have this option back with a message that basically says "use at your own risk".

Can we get that official response please?

Link to comment
Share on other sites

I'm posting just to bump this, I'm big fan of affinity as an adobe alternative but shooting yourself in the foot like this is ridiculous, just allow user to enable it and test it out with default being off.

 

Also, why don't you offer paid major feature upgrades, maybe once in two years? With the speed and size of the updates it feels like you're struggling, I would gladly pay for proper new stable major version, I wonder if the current way is sustainable.

Edited by mb69
Link to comment
Share on other sites

Seeing as everyone else is piling in here, I thought I would too.  @omiK I agree. Some clarification at least would be useful and much appreciated.  Perhaps a mooted v2 is what's stopping updates here?

As a side note, in the latest release notes from AMD (22.6.1) there is a somewhat oblique reference to OpenCL performance with their drivers, viz

"Lower than expected Folding@home™ compute performance with OpenCL™ API on some AMD Graphics Products such as the Radeon™ RX 6800"

Although this impacts folding@home I wonder if it's related to the problem Serif reported.  Maybe it means the the issue is (finally) bubbling its way to the surface as far as AMD are concerned.

Link to comment
Share on other sites

38 minutes ago, Subclavius said:

Seeing as everyone else is piling in here, I thought I would too.  @omiK I agree. Some clarification at least would be useful and much appreciated.  Perhaps a mooted v2 is what's stopping updates here?

As a side note, in the latest release notes from AMD (22.6.1) there is a somewhat oblique reference to OpenCL performance with their drivers, viz

"Lower than expected Folding@home™ compute performance with OpenCL™ API on some AMD Graphics Products such as the Radeon™ RX 6800"

Although this impacts folding@home I wonder if it's related to the problem Serif reported.  Maybe it means the the issue is (finally) bubbling its way to the surface as far as AMD are concerned.

thx for checking that out, looks like there may still be hope for this issue

Link to comment
Share on other sites

  • 2 weeks later...

Has anybody found a way of leaving a message on an AMD site re this issue - I have visited their site and, like most companies, they seem to send you in loops circles and get nowhere - perhaps there is some AMD forum?

Mel

Link to comment
Share on other sites

46 minutes ago, Mel_P said:

Has anybody found a way of leaving a message on an AMD site re this issue - I have visited their site and, like most companies, they seem to send you in loops circles and get nowhere - perhaps there is some AMD forum?

Mel

Serif need to be the ones doing this. Individuals will have no effect at all unless there was a huge number of us talking about a blockbuster game title.

Link to comment
Share on other sites

52 minutes ago, Mel_P said:

Has anybody found a way of leaving a message on an AMD site re this issue - I have visited their site and, like most companies, they seem to send you in loops circles and get nowhere - perhaps there is some AMD forum?

Mel

Open Radeon software and find the bug icon (looks like a lady bug) press to report a bug.

Current Workstation:
CPU: AMD Ryzen 5 5500 - MOBO: Asus B450 - RAM: 16GB DDR4 2667Mhz - GPU: AMD Radeon 7850 1GB
NVMe SSD: Crusial P3 1TB M.2 -  SSD: Samsung Evo 850 256GB  - PSU: XFX TS450 - OS: Win10

Link to comment
Share on other sites

by the way davinci resolve is apparently using opencl for its gpu accel on my 6600 card and it is running very well... I would pay 20 usd no problem for proper gpu accel for affinity. But I guess they have the stats and amd card users are probably such a small percentage of the userbase, multiplied by small percentage that have heavy enough use cases to even care about the gpu accel, that they decided not to bother. Back to photoshop I guess, shame, it really felt great with the GPU, not even the big operations but the general responsivness of the layer stack gui etc...

Edited by mb69
Link to comment
Share on other sites

Since this is an ongoing issue, Serif should state that support for Radeon Graphics cards is limited and they recommend nVidia GPU. I just bought Radeon 6700 XT and after installing it and discovering this problem, I feel like an idiot. Serif should be the one to deal with AMD, not the users.

Link to comment
Share on other sites

4 hours ago, ysbuff said:

Since this is an ongoing issue, Serif should state that support for Radeon Graphics cards is limited and they recommend nVidia GPU. I just bought Radeon 6700 XT and after installing it and discovering this problem, I feel like an idiot. Serif should be the one to deal with AMD, not the users.

I have an RX 6600 - similarily totally hacked off!

Link to comment
Share on other sites

I just bought a RX 6800 XT and finally switched from my 1060, a purchase for which I have saved money for over 2 years. I'm experiencing worse performance in my project files than before, now I'm here, read the thread and my disappointment is unmeasurable...

Link to comment
Share on other sites

13 hours ago, Nox said:

I just bought a RX 6800 XT and finally switched from my 1060, a purchase for which I have saved money for over 2 years. I'm experiencing worse performance in my project files than before, now I'm here, read the thread and my disappointment is unmeasurable...

I though about having 2 different GPUs to avoid this stupid problem in my other system where I have the 5500XT, I was told that it is possible to have an nvidia alongside the amd GPU and check in affinity to use the nvidia one.

Stupid? yeah, but it can be a temporary solution till serif decide to tell us more on the topic or fix it with amd, if you still have the 1060 around.

Current Workstation:
CPU: AMD Ryzen 5 5500 - MOBO: Asus B450 - RAM: 16GB DDR4 2667Mhz - GPU: AMD Radeon 7850 1GB
NVMe SSD: Crusial P3 1TB M.2 -  SSD: Samsung Evo 850 256GB  - PSU: XFX TS450 - OS: Win10

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...

Important Information

Terms of Use | Privacy Policy | Guidelines | We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.