If you could change one thing about GPU architecture for LLMs, what would it be, and what argues against it?

A proposal for more memory bandwidth needs a workload that benefits, a budget for what it costs, and a reason to abandon it. Work through a serving-specific design without assuming all LLM work has the same bottleneck.

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