ByteBrief
We're a portrait publication through and through. Turn your phone back and your briefing picks up right where you left it.
(We tried widescreen once. It wasn't us.)
Meta's MTIA 400 accelerator targets LLM training and ad recommender inference. It delivers 12 petaFLOPS, about 20 percent faster than Nvidia's top Blackwell parts, but is 3x to 3.3x slower than Rubin and AMD's MI455X. The MTIA 450 and 500 are planned.
Tap to vote and see what everyone thinks.
Summary by ByteBrief