Different levels of performance from same GPU types in Community Cloud
When I use an A5000 in Community Cloud, I am able to get over 3it/s training Kohya_ss in ES and FR regions, but only a pathetic 1s/it in BG region. If different hosts are going to offer different levels of performance, they should not all earn the same fixed rate. A less performant host needs to earn less than the ones that have decent performance.
11 Replies
Both ES and FR go up to about 200/230W when training is in progress, but the nerfed piece of shit BG host only goes up to 88W/230W. The host seems to have some kind of power cap enabled even though its not showing in
nvidia-smi. This is unacceptable and this host needs to be removed.Do you mind running my tester tool and sending output on DM
I'll do it when that garbage host ios working again.
I hate autocorrect
Do you at least have pod id of that pod?
It shows A5000 available but then I get this, this host well and truly is the worst host on RunPod.
Huh would need to find what is machine is for that need either script or pod id
07ibexeyrwp5
Why can't we copy and paste the pod id from here? 🤦♂️
Pretty stupid having audit logs we can't copy and paste from.
Got it forwarded to team and asked for permission to delist that machine
07ibexeyrwp5 is actually the machine id. Audit logs are showing the machine id for some reason instead of pod id.
Actual pod id: crvugt7qnz0ye5.
Every time I am able to get a an A5000 pod in BG, its always that same machine id.