causal
System Efficiency Explains Gap Between GPU and Cluster
Even though a single Nvidia GPU runs at 83 percent efficiency, cluster-level MFU falls to about 35 to 40 percent because system FLOPS efficiency, covering networking, storage, and memory, captures the losses that happen as data moves to the GPU.
causalpending
Speaker
GavinEvidence Quote
“And the reason is something I would call SFU, system FLOPS efficiency. And this really captures networking, storage and memory”
Source
Gavin Baker - AI, Semiconductors, and the Robotic Frontier - [Invest Like the Best, EP.385]— Invest Like the BestCreated: 11/12/2026, 2:50:15 AM
Your notes
Loading notes...