ML4 Trained on 3,800 Grace Blackwell GPUs; Mistral's C and D Round Clusters Coming Online
GuillaumeLample · x · 2026-10-06
Part 2 of the ML4 thread: the model was trained on 3,800 NVIDIA Grace Blackwell GPUs at Mistral's Bruyères-le-Châtel datacenter in Europe, built with Series B funds. Series C and D clusters are coming online, enabling longer training, more ambitious post-training, and faster iteration, with large and rapid improvements expected in the weeks and months ahead.
Related event: Mistral Launches Mistral Large 4, a 1T-Parameter Open-Weight Flagship(86 posts)→
More from Infra
- Sarvam resells DeepSeek in India, drawing backlash over AI sovereignty stance — teortaxesTex · 2026-11-21
- Ilya Sutskever warns weak neocloud security could let rogue agents hijack compute — npinto · 2026-11-21
- Colfax and MIT speed up lattice QCD staple networks 700–1200x on H200 with CuTe DSL — snowclipsed · 2026-11-21
- Will MoE Inkling 2 Mini get strata support and rival dense Qwen on 12GB VRAM? — ECrispy · 2026-11-21
- 'Computers Have Broken Everything': NYT on America's growing tech fragility — nordicinst · 2026-11-21
- Can foundries build photonic processors with unclonable fingerprints? — jwt0625 · 2026-11-21