| name | gpu-sku-selection |
| description | Pick the right N-series SKU (NC, ND, NV) for training / inference / visualization |
Gpu Sku Selection
When to use this skill
- Workload uses GPU acceleration
Key concepts
- NC — compute training/inference; v3 V100, A100, A10; H100 in H100 v5.
- ND — large model training; A100 80GB, H100/H200; NVLink + InfiniBand.
- NV — visualization/VDI; A10, T4.
- ND MI300X v5 — AMD Instinct MI300X for large model training.
- Memory matters: pick by model size (e.g., 70B FP16 ≈ 140GB → A100 80GB pair).
Common pitfalls
- Using NC for visualization — drivers and license profile mismatch.
References
Validation-first: verify every Azure fact against the Microsoft Learn MCP server (cite the Learn URL); Learn overrides built-in knowledge. If no Learn MCP server is configured, flag the answer ⚠️ unverified and mark specs as indicative. Analysis only — verify against Microsoft documentation before applying.