Reputation: 1
My request for a quota increase for Nvidia A100 GPUs was approved. But for some stage reason, it is not reflected in the AI platform machines, while using notebooks.
In 1 we see that the quota increase was approved for me for europe-west4, but the machine region is not available at my AI platform instances 2.
Is there an additional intermediate step missing in order to use the device?
Upvotes: 0
Views: 233
Reputation: 10058
Try this:
gcloud beta compute instances create a100-test1 \
--zone=europe-west4-a \
--machine-type=a2-megagpu-16g \
--subnet=default \
--network-tier=PREMIUM \
--maintenance-policy=TERMINATE \
--scopes=https://www.googleapis.com/auth/cloud-platform \
--accelerator=type=nvidia-tesla-a100,count=16 \
--image-family=tf-latest-gpu \
--image-project=deeplearning-platform-release \
--metadata="proxy-mode=mail,[email protected]"
You can try different A2 model/image family/A100 cores
Upvotes: 0
Reputation: 4051
Currently, AI Notebooks are not available in europe-west4. You can confirm it by using the following command,
gcloud beta notebooks locations list
You will see that europe-west4 is not in the list.
However, there is an open Feature Request to add this region, here. You can go to the page, click on +1 to indicate it affects you and post a comment there. Then follow the issue tracker case for any updates.
Upvotes: 1