Runpod•3mo ago

No workers available in EU-SE-1 (AMPERE_48)

I deployed endpoint

s7gvo0eievlib3

s7gvo0eievlib3

s7gvo0eievlib3

s7gvo0eievlib3 hours ago with storage attached. Build was fine and release was created. But I don't have any workers assigned. The GPU is set to

AMPERE_48

AMPERE_48

AMPERE_48

AMPERE_48 of which it said High Supply. What am I doing wrong and how do I fix this?

Dj•9/29/25, 4:43 PM

Without looking, it may be based on your CUDA version filtering - but I'll take a look for ya.

Dj•9/29/25, 4:44 PM

I'm in a meeting so I can't get you the specific versions just yet but you should be a lot better enabling the CUDA 12.9 version as well. There's backwards compat :)

JanOP•9/29/25, 4:46 PM

I have 12.8 as min requirement. Just switched everything to EU-IS-1 to see if I can get RTX 5090 assigned there.

JanOP•9/29/25, 4:49 PM

Same in EU-IS-1, I'm not getting workers assigned

Dj•9/29/25, 5:02 PM

It's not min requirement at the moment, its a hard filter. You will need to manually also select 12.9.

Dj•9/29/25, 5:02 PM

Like I said I can check exact CUDA versions in a little bit here to see exactly what's going on I'm multitasking a meeting

JanOP•9/30/25, 5:40 AM

I've switched allowedCudaVersionsallowedCudaVersions to 12.8,12.912.8,12.9 yesterday, but not a single worker showed up to work today.

JanOP•9/30/25, 5:40 AM

JanOP•9/30/25, 5:40 AM

JanOP•9/30/25, 8:43 AM

Is there a global limit on the number of workers?!

JanOP•9/30/25, 10:44 AM

Closed, solved. Thank you!

JanOP•9/30/25, 10:44 AM

Issue was it seems I hit a global worker limit which I increased, but the serverless endpoint had workers set to 0

No workers available in EU-SE-1 (AMPERE_48)

Similar Threads

Similar Threads

Similar Threads