Lab 18: Noisy Neighbors on a Shared GPU with HAMi and Linkerd
Two physical GPUs, one cluster, twelve proofs. What HAMi isolates, what it can't, and how a service mesh sees the difference and routes around it.
Two physical GPUs, one cluster, twelve proofs. What HAMi isolates, what it can't, and how a service mesh sees the difference and routes around it.
Run multiple Pods on one GPU with enforced VRAM and compute limits.
Share one non-MIG GPU between Pods on single-node k3s and prove HAMi-core enforces the memory cap.