I submitted an AWS quota increase request on September 11 for a fairly modest GPU workload. I've followed up twice, but the request is still unassigned and I haven't received a response.
The account is more than six years old and in good standing. I'm trying to migrate some workloads from Vultr after they discontinued fractional GPUs, so I was hoping to run them on AWS instead.
Is this kind of delay normal for GPU quota requests, or is there a better way to get the request reviewed?
4 Answers
GPU availability is a major factor. If possible, check other AWS Regions and submit requests where the instance type you need has better capacity. You may have more luck spreading the migration across multiple Regions.
You could temporarily enable Business Support and submit the request again or ask support to review the existing case. Just remember that the plan is billed for a full month, so it makes sense to prepare a list of all the quota increases you may need before subscribing.
Unfortunately, that can happen with GPU requests, especially if the account doesn’t have a paid support plan. GPU capacity is tightly managed, so AWS may review the request manually rather than approving it automatically.
Include a clear, human-written note in the quota case explaining the workload, instance types, expected usage, migration plan, and estimated growth. Quota requests often create a support case automatically, and adding that context can make the eventual review easier.
That’s useful to know. I’ll gather the other quota requests first so I’m not paying for another month just to submit one case.