You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I'm writing matmul kernels right now, and I have an issue where I will run my kernel 100 times and each time get drastically different performance. I suspect this is caused by macos shifting the gpu to low power mode / limiting it's clock speed.
There's also the issue where when I switch windows during execution, my kernels start returning 0's and taking 0 time to run.
Has anyone dealt with this, or is there a ggml issue I can look at where similar problems have been addressed?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
Hi, I'm writing matmul kernels right now, and I have an issue where I will run my kernel 100 times and each time get drastically different performance. I suspect this is caused by macos shifting the gpu to low power mode / limiting it's clock speed.
There's also the issue where when I switch windows during execution, my kernels start returning 0's and taking 0 time to run.
Has anyone dealt with this, or is there a ggml issue I can look at where similar problems have been addressed?
Beta Was this translation helpful? Give feedback.
All reactions