1 min read
2.1 · One GPU, Many Jobs: The Case for Sharing

Series stub — full post TBD. This page exists so the series shape is reviewable.

Planned focus: A whole accelerator handed to a job that uses a sliver of it is the most common waste; the fork between splitting in space and sharing in time.


Part of “Inside AI Infrastructure: The Compute Layer.” Opinions are my own; public, documented concepts only.