Design the complete network for a 1,024-GPU cluster: compute fabric, storage, management. What are the counts and where does the money go?

Derive a complete rail topology and distinguish physical links, switch ports and transceivers. Size storage and management from their own requirements.

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

Derive a complete rail topology and distinguish physical links, switch ports and transceivers. Size storage and management from their own requirements.

This answer is included with a free account. Sign in to read it and save your progress.

free account · no card needed

The concepts behind this question

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

Foundational

🖧 Hardware & Cluster Build-Out

The Bill of Materials for a Training ClusterA cluster bill of materials includes networking, storage, management, power, cooling and spares as well as GPUs. Size each from the workload and topology.

Advanced

📐 AI Systems Design🔒 Premium

Training Cluster Design at 10k GPUsSize a training cluster from the model, data and completion deadline. Then check the network, storage, power and recovery capacity needed to keep those GPUs productive.

Advanced

📐 AI Systems Design🔒 Premium

Model Routing and CascadesChoose a model against measured task quality, latency and cost. Distinguish routing before generation from cascades that pay for a first answer before escalating.

Advanced

🔌 Networking & Storage🔒 Premium

Rail-Optimized and Fat-Tree FabricsRail-optimized and Clos fabrics organize traffic between GPU nodes. Port counts, placement and oversubscription determine which paths are available to collectives.

Explore the full curriculum →

UP NEXT ON YOUR JOURNEY

Next in this trackDoes multi-node inference need the same fabric as training, or can it run on ordinary networking?Next in this trackOne fabric link goes down and comes back once an hour. What does that do to a 1,024-GPU training job, and how would you find it?Next in this trackWhy does object storage behave badly for random reads, and what does a training pipeline do about it?

FEDITOR'S NOTE

Require a consistent port map and cost boundary. Test shared dependencies rather than promising that separate networks eliminate every source of interference.

DISCUSSION · 0

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