Skip to main content

Mainboard vs GFX processing - is GFX processing always better?



  • Mark Suazo

    Follow up -

    The card is built with the DDR5X while the memory on the mobo is DDR4. In my case, following the "canned" advice would reduce the amount of RAM the application can utilize by 50%ish (based on total load) (16GB rather than 32GB). But the GFX card has the CUDA cores... and my google machine says DDR5 is 2x faster than DDR4.
    As I understand it, I can either run On1 "stock" - which I believe would utilize the system memory - or I can set it up to run exclusively off of the GFX card via the NVIDIA control panel. Hope to know which is most effective without trying benchmark it myself. With 32GB onboard, I'm not all that concerned about reducing the load on the system RAM. (And I'm not easily offended, so if this is a dumb question, feel free to let me know. :-)
  • Nitin Chandra

    I guess only ON1 can answer that besides someone who might take the time to extensively test it out. The one thing I can tell you is that instead of using RAM, ON1 operations like effects etc are all passed around on disk and having your ON1 scratch space on a fast SSD would help.

  • Brian Lawson Community moderator

    GPUs are faster at processing graphics that a CPU. For best performance you want the program to make as much use of the GPU as possible. You're already noted that the VRAM is DDR5 while your system RAM is DD4 so you can see the video card memory is faster.


Please sign in to leave a comment.