Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
loading thread…
3 ms
·
OpenAI GPT-OSS models use MXFP4 to cut inference costs
8 points
by
rntn
1y ago