Runpod underwater?
Hey everyone,
First message, prompted by what I've been seeing for a while at the website.
I'm launching a new project soon and I've been tracking availability of some GPUs in Europe. It suddenly went through the floor! EU-RO-1 disappeared altogether, for instance...
Posting it here, by the way, because the serverless facility is (was?) the critical factor in choosing Runpod.
Quite disconcerting to read all the serverless/workers problems people are encountering... Worse still, that these problems seem to be persisting over days on end!
So my question to Runpod: These do not seem to be isolated incidents, so what's really going on?
4 Replies
@ST.PT.LV We are all waiting for a statement i guess. I talked to runpod support and they gave me credits back for stuck workers that i got charged for, however they didnt really give any informations about the availability / initialization issues.
Unknown User•2d ago
Message Not Public
Sign In & Join Server To View
That's my point. Try finding, for instance, an A100 80GB available anywhere in Europe. From memory, perhaps only and very limited in... Iceland! Clearly, Runpod has a lot of work to do to boost availability.
Further, these ongoing issues with the serverless endpoints, their key differentiator, are a major red flag for me. What good is securing GPU/storage, if afterwards the service isn't working or dependable?
Unknown User•2d ago
Message Not Public
Sign In & Join Server To View