Jump to content
Mark Ingram

Benchmark 1900 Results

Recommended Posts

8 minutes ago, Chris B said:

Wow thanks Max N, that's very much appreciated! Glad to know there's a bit of a boost as well as hopefully stability etc. 

I think that in real use the result may differ from the benchmark. And I hope that for the better.

By the way, an error in translation (chenchmark as a checkpoint) remained in RC. Hopefully they will fix it for the final release.

 


__________________

Windows 10 64-bit,

AMD Ryzen 5 3600 + Nvidia 1660 Super + Nvidia Studio driver + 32 Gb RAM.

Share this post


Link to post
Share on other sites

W10, Ryzen 7 1800X, 32GB, Nvidia GTX1080

Unfortunately feels slower with OpenCL active in most situations.  


 

image.png

Share this post


Link to post
Share on other sites
1 hour ago, NotMyFault said:

W10, Ryzen 7 1800X, 32GB, Nvidia GTX1080

Unfortunately feels slower with OpenCL active in most situations.  


 

image.png

That's interesting - we've heard similar complaints from other owners of AMD Ryzen machines. I'm glad you've provided a benchmark though, because in theory, your GPU should be ~25X faster than your CPU. What kinds of workflows are you noticing the penalty with?

Share this post


Link to post
Share on other sites
21 hours ago, Mark Ingram said:

That's interesting - we've heard similar complaints from other owners of AMD Ryzen machines. I'm glad you've provided a benchmark though, because in theory, your GPU should be ~25X faster than your CPU. What kinds of workflows are you noticing the penalty with?

Hi Mark, @Mark Ingram

 

i tried to attached a sample file which became uneditable., but upload always failes around 200 MB (and not 1140,45MB). with a message box "Error -200".

image.thumb.png.ff32e5f95ec0b608f6e1d8e431ef910a.png

 

Share this post


Link to post
Share on other sites
34 minutes ago, NotMyFault said:

Hi Mark, @Mark Ingram

 

i tried to attached a sample file which became uneditable., but upload always failes around 200 MB (and not 1140,45MB). with a message box "Error -200".

 

I think there’s an upload limit on the forum. Can you upload to Dropbox or similar?

Share this post


Link to post
Share on other sites

AMD Threadripper 3970X 32-core 3.70/4.50 Ghz
MSI RTX 3090 Suprim X 24GB + Nvidia studio driver 461.40
64GB RAM @ 3800mhz CL14
Windows 10 Pro, Version 2004

Any idea why my GPU results are 10% slower than RTX 3080?

affinity-photo-1.9-RTX3090.png.875e38bb2e47a013b3ffe1fdba4879a3.png


Affinity Photo 1.9.0.932 for Windows  OS: Windows 10 x64 2004  CPU: AMD Threadripper 3970x  RAM: 64 GB 3800 Mhz  GPU: MSI GeForce RTX 3090 Suprim X 24GB + studio driver 460.89  NVMe SSD Samsung 960 Pro 512 GB  Monitor: Eizo ColorEdge CS2420 24"

Share this post


Link to post
Share on other sites
5 hours ago, Alex_M said:

AMD Threadripper 3970X 32-core 3.70/4.50 Ghz
MSI RTX 3090 Suprim X 24GB + Nvidia studio driver 461.40
64GB RAM @ 3800mhz CL14
Windows 10 Pro, Version 2004

Any idea why my GPU results are 10% slower than RTX 3080?

Try it a few times, there can be quite a variation in GPU numbers.

Share this post


Link to post
Share on other sites

Perhaps it is just me, but the benchmark results are pretty abstract to me. What does it mean in the end? How fast is a job done w/wo hardware acceleration? I don't know if this is doable: A reference image and a reference macro doing some heavy operations on the image. So with a stop watch one could measure the difference. Sounds silly. 🙄


------
Windows 10 | i5-8500 CPU | Intel UHD 630 Graphics | 8 GB RAM | Latest Retail and Beta versions of complete Affinity range installed

Share this post


Link to post
Share on other sites
On 2/7/2021 at 1:47 AM, Mark Ingram said:

That's interesting - we've heard similar complaints from other owners of AMD Ryzen machines. I'm glad you've provided a benchmark though, because in theory, your GPU should be ~25X faster than your CPU. What kinds of workflows are you noticing the penalty with?

I have the same impression with our video workstation (mediocre hardware: Ryzen 7 3700X + Radeon RX 5700XT + 64GB): With hardware acceleration switched on everthying feels a bit laggy. When I used "File -> New from Clipboard" for the attached screenshot, it took 3 seconds to get displayed. And when I applied the crop it stuttered again. Then I performed the same tasks with acceleration switched off - and everthing was butter smooth.

Funny: The Intel-powered laptops in our office (although much slower in general) don't show such a behavior.

 

benchmark.jpg

Share this post


Link to post
Share on other sites
1 minute ago, VolkerMB said:

I have the same impression with our video workstation (mediocre hardware: Ryzen 7 3700X + Radeon RX 5700XT + 64GB): With hardware acceleration switched on everthying feels a bit laggy. When I used "File -> New from Clipboard" for the attached screenshot, it took 3 seconds to get displayed. And when I applied the crop it stuttered again. Then I performed the same tasks with acceleration switched off - and everthing was butter smooth.

Funny: The Intel-powered laptops in our office (although much slower in general) don't show such a behavior.

Thanks @VolkerMB. We've identified a problem with the Radeon RX 5000 series and greater (NAVI architecture). Could you post your findings into this thread for us please?

 

Share this post


Link to post
Share on other sites
14 hours ago, Mark Ingram said:

Try it a few times, there can be quite a variation in GPU numbers.

I tried a dozen times. It doesn't go any higher than the score I posted.


Affinity Photo 1.9.0.932 for Windows  OS: Windows 10 x64 2004  CPU: AMD Threadripper 3970x  RAM: 64 GB 3800 Mhz  GPU: MSI GeForce RTX 3090 Suprim X 24GB + studio driver 460.89  NVMe SSD Samsung 960 Pro 512 GB  Monitor: Eizo ColorEdge CS2420 24"

Share this post


Link to post
Share on other sites

Affinity Photo 1.9.1.  Mid 2014 MacBook Pro with a Nvidia GT 750M GPU which seems to be ignored even though it claims to support Metal.  Or is it an Affinity Photo bug?

1644094422_ScreenShot2021-02-26at9_37_30PM.png.f7cf7790ea8d8fa8927e4a8fd8e27aa2.png

 

2030462026_ScreenShot2021-02-26at9_37_53PM.png.7c592380143b1c925479645f17cb4a48.png

Edited by Steve N

Share this post


Link to post
Share on other sites

Please can all readers of this thread use the latest beta build and repost in this new thread, as the algorithm and test has been changed and the new results would be very interesting to us.

 This thread is now locked


Patrick Connor
Serif Europe Ltd

Latest releases on each platform 

There is nothing noble in being superior to your fellow man. True nobility lies in being superior to your previous self."  W. L. Sheldon

 

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.

×
×
  • Create New...

Important Information

Please note there is currently a delay in replying to some post. See pinned thread in the Questions forum. These are the Terms of Use you will be asked to agree to if you join the forum. | 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.