animetosho
commented
3 weeks ago ParPar has experimental OpenCL support. It's disabled by default and you have to opt into it, as well as manually allocate a processing portion.
As CUDA is Nvidia specific, it's unlikely I'll support it any time soon.
Note that GPU performance seems to be much worse than CPU, so enabling OpenCL may not improve performance much over just CPU.
Parpar performs very quickly with small file sizes but is slower with larger files, such as 50 GB RARs, compared to other tool . Multipar with max options.
You'll need to share much more info if you want input from me, including your hardware specifications, exact settings used, the files used (one 50GB RAR file? multiple 50GB RAR files? some split RAR? what?) and other relevant info, including the console output of the applications etc.
Parpar GUI is slow ?
I presume you mean just ParPar?
Or do you mean the GUI itself is slow? (i.e. slow to react to user input?)
Hello, Could you please add support for multiple CUDA-enabled GPUs in the PARPAR tool to fully utilize high-end GPUs alongside CPU, RAM, NVMe, and GPUs with CUDA? Parpar performs very quickly with small file sizes but is slower with larger files, such as 50 GB RARs, compared to other tool . Multipar with max options.