A team wants NVIDIA software on a cloud GPU instance, with little setup. What should it deploy?
Why C is the answer
An NVIDIA VMI
VMIs are preconfigured virtual machine images.
In detail
- VMIs are preconfigured virtual machine images.
- VMIs are preconfigured.
Use case
An illustrative scenario, not a real incident.
A team starts a cloud GPU instance from an NVIDIA VMI, with the software already there.
See it in different ways
flowchart LR A["NGC catalog"] --> B["API key if needed"] B --> C["Pull from nvcr.io"] C --> D["Run with GPUs"]
sequenceDiagram participant U as Operator participant R as nvcr.io participant H as GPU host U->>R: Signs in with an NGC API key R-->>H: GPU-optimized image H->>H: Runs with GPUs
flowchart LR S0["Choose VMI"] --> S1["Launch instance"] --> S2["Run work"]
Key point: VMIs reduce cloud setup.
Why the other options are wrong
- A. An InfiniBand subnet manager
- That runs an InfiniBand fabric.
- B. A BMC recovery image
- That repairs a BMC.
- D. A DGX firmware bundle
- That is for DGX firmware.