The cluster dashboard says 90% allocated and 30% utilized. What is happening, how do you prove it, and what policy fixes it?

First align the time window and denominator. Then attribute allocated capacity, kernel activity and useful progress to jobs before choosing between reclaim, queue policy and performance work.

Updated Oct 2026 · Learn how AI infrastructure works through explanations, worked examples and diagrams, then practise applying it to interview questions.

First align the time window and denominator. Then attribute allocated capacity, kernel activity and useful progress to jobs before choosing between reclaim, queue policy and performance work.

The full answer is part of Premium. A free account includes more answers in each topic, but does not unlock this one.

20 answers per topic instead of 10, plus saved progress and bookmarks · no cardor unlock all 316 remaining answers · ₹2,000 / $25

The concepts behind this question

Use these concept explanations to review the ideas behind this question.

Foundational

🩺 Fleet Reliability & Observability

Alert Design and On-Call LoadUseful alerts identify a condition that needs a timely response. Connect pages to user impact or imminent risk, and reduce duplicate noise without hiding failures.

Advanced

🗂️ Scheduling & Orchestration🔒 Premium

Multi-Tenancy, Quotas and Fair ShareFollow three teams through GPU borrowing and reclaim. Distinguish quotas, reservations and fair sharing; see why a returning team may leave another above its nominal share, and why failures or indivisible jobs change what can be promised.

Advanced

🗂️ Scheduling & Orchestration🔒 Premium

Gang Scheduling with Kueue and VolcanoGang scheduling treats a distributed job’s required workers as one admission unit. It prevents partial allocations from holding resources while waiting for the rest of the job.

Core

🗂️ Scheduling & OrchestrationSign in

MIG, MPS and Time-SlicingMIG, MPS and time-slicing share a GPU in different ways. Compare memory capacity, execution behavior and isolation before choosing a mechanism.

Explore the full curriculum →

UP NEXT ON YOUR JOURNEY

Next in this trackWe have GPUs in two clouds. Can we run one training job across both, and what would you do instead?Next in this trackDesign a notebook platform for 300 researchers on 64 GPUs. How do you share, reclaim and account for the GPUs?Next in this trackHow do you give a Kubernetes pod RDMA access to the fabric for NCCL, and what goes wrong when it is done naively?

FEDITOR'S NOTE

Treat allocation, activity and progress as separate measurements. Do not turn fleet averages into a count of reclaimable cards or identify a bottleneck from tensor activity alone.

DISCUSSION · 0

No comments yet — be the first to share your approach.