C1 20 GPU acceleration uses Intel GPU and NOT Nvidia GPU on laptop
I have tried un-installing C1, disabling the Intel GPU so the only option is the Nvidia card, re-installing C1 and it will say that hardware acceleration is enabled, but resource monitor shows it's only using CPU for all tasks. Nvidia card gets very little utilization.
If I re-enable my Intel GPU it will utilize that one to 100% when doing things like a batch process, but, that GPU is even slower than my CPU!
It looks like others have talked about having similar issues... has there been a fix yet? Anyone know a work-around so that it can be forced to use my Nvidia GPU? It's the MX230, if that matters at all.
-
Billy,
There's a Log File for the GPU driver assessment here
C:\Users\[USER NAME]\AppData\Local\CaptureOne\Logs\ImageCore.log
This records what C1 found to be available GPUs and how it assessed them.
It's not the clearest of files in that the assessments are run in parallel and the lines reported as the results become "known" - so they can be intermingled. Have a look to see what you can make of it.
If log files are not your thing if you can place it somewhere secure but shareable there are people here who should be able to give you some guidance about what it is reporting.
In my experience, but with older hardware I suspect, it is never the CPU based GPU that is used although in theory C1 will use a number of GPUs at the same time.
Grant
0 -
Same behavior here. Frustrating.
Not super current (to C1v12), but a relevant discussion on C1 GPU performance findings.
Is there a way to force C1 to re-evaluate the hardware acceleration?
0 -
Same on my Acer Aspire with a Nvidia GTX 960M with four gigs of ram. Capture One utilizes 0% of it. No ram no processing.
0 -
As mentioned above, there is a log file dedicated to recording the activities of the GPU device discovery and assessment process that runs, at least as a check for configuration changes, every time you start Capture One (assuming that you are using V20.
The first step is to understand C1 sees of your installation.
Also check the Preferences > General tab to make sure that the Hardware Acceleration options are set to "Auto" and the status message indicates that Hardware acceleration is active.
You can also check the Nvidia monitor to see which programs it thinks are actively being enabled.
If everything suggests that the Nvidia card is considered to be active it is then a matter of looking at the processing you are doing (not all processing benefits from the GPU approach) and what you are expecting to see from whatever monitor you are using. There may also be a need to check the NVidia configuration to see if there are any unusual settings applied.
0 -
Same with C1 12 Fujifilm (Windows) on Lenovo P50s with discrete Nvidia Quadro M500M GPU. Frustrating...
0 -
Not necessarily frustrating Vitali.
My older notebook has an earlier Quadra of similar overall specification and performance, although a few different build in code support options it seems.
The potential performance it offered was only just considered useful according to the Hardware assessment log file. In practical use, I could rarely notice any difference in performance whether I had it in use or turned off. That was for a 2012 reasonably high spec machine.
My 2020 specification equivalent (in terms of market positioning) notebook can be made to show a much more identifiable benefit in performance with Hardware Acceleration activated - but then it can also make use of the software based Intel GPU which was never possible in the old CPU.
You may not be missing as much as you think - but check the assessment logs as mentioned in the post above.
0 -
Maybe you are right SFA - right now C1 works more or less acceptable with CPU only, but I find it difficult to evaluate this, since there is nothing to compare with. When I was a Canon user, I've used DxO Optics Pro, and AFAIK it did support GPU on this the very same laptop.
0 -
GPU memory of you Nvidia is too smal or is almost 100% full. Than C1 to start using GPU from Intel or CPU.
0 -
I know, it is an old thread but I've been stumbling across it every time I was looking for an solution.
I was having the same problems and it was driving me crazy. Processing files in C1 was utilizing my internal GPU up to 100% but my dedicated NVIDIA GPU only up to 3% (up to 17% with a previous version of C1).
I changed the following setting in Windows 11 and now it's using both GPU's up to 100%.
System -> Display -> Related settings -> Graphics -> Default settings -> Change default graphics settings ->"Hardware-accelerated GPU scheduling" set to "On "
But in case it does work for you you still need to check whether it really increases the performance on your hardware. For me, it did not. Even though it was now using both GPU's (also the faster one, which it did not before), exporting images was actually slower than it was before.
Exporting 245 images with Hardware-accelerated GPU scheduling ON took 5 minutes 50 seconds
Exporting the same 245 images with Hardware-accelerated GPU scheduling OFF took 4 minutes 52 seconds
0
Post is closed for comments.
Comments
9 comments