Maia 200 hits ~12 TFLOP/s FP4 in 1mm²: density should be a first-class goal
thoefler · x · 2026-09-14
The author argues the industry should discuss flop/s per square millimeter, not just flop/s: Microsoft's Maia 200 delivers 12 TFLOP/s FP4 (6 TFLOP/s FP8) in just 1 mm²—roughly a grain of salt. When data movement dominates energy consumption, density becomes a first-class design objective.
More from Infra
- Rumble's Neocloud Pivot Lands $13.7B Compute Deal With Anthropic — andersonbcdefg · 2026-09-14
- Running an LLM agent on a 512MB board with decoupled memory and live cross-machine migration — D777Castle · 2026-09-14
- Dev's 24/7 self-hosted AI stack: OpenWebUI, pidot, Tailscale, GLM and DeepSeek — andfanilo · 2026-09-14
- Dream Photonics' laser integration render caught mirroring the whole chip image — jwt0625 · 2026-09-14
- Engineer questions the inline PTX hype DeepSeek sparked: hand-written PTX isn't a proxy for performance — mike64_t · 2026-09-14
- Qwen3.8-Flash-Next on M3 Ultra: 559 t/s prompt processing, 31 t/s generation — rm-rf-rm · 2026-09-14