Standard_NC40ads_H100_v5 AML Compute Cluster Node has only 128 GB of storage #3749
Description
Operating System
Windows
Version Information
I am creating a Cluster and a training job via Azure ML Studio
Steps to reproduce
According to the documentation (https://learn.microsoft.com/en-us/azure/virtual-machines/ncads-h100-v5) the Standard_NC40ads_H100_v5 VM of the h100_v5 compute familty has 3576 GiB of storage, but a compute cluster node of that vm size has only 128 GB of memory. This goes against the purpose of the compute (like training LLM) because bigger-than-small LLM wont fit inside the disk to even start training.
Expected behavior
The cluster node should have around 3576 GiB of disk storage as advertised.
Actual behavior
The node has only 128 GB of storage so it's not possible to use it to train or run even medium models (like Qwen3-Coder-Next)
Addition information
My workspace is in the West Europe region.