normative
NVLink Mandatory for Low-Latency, Not for Sail
NVLink is mandatory for low-latency inference and tensor parallelism is basically Nvidia-only, so Sail does not care much about low latency and instead optimizes flops per dollar using other vendors' chips with alternative parallelism such as expert or pipeline parallelism.
normativepending
Speaker
Neil MovvaEvidence Quote
“how many flops does this chip have and how much is it going to cost me per hour to own and operate?”
Created: 11/17/2026, 6:43:49 AM
Your notes
Loading notes...