Mistral trained ML4 on 3,800 Grace Blackwell GPUs in its European datacenter, more clusters coming
rakyll · x · 2026-10-07
Mistral co-founder Guillaume Lample revealed that ML4 was trained on 3,800 NVIDIA Grace Blackwell GPUs at the company's European datacenter in Bruyères-le-Châtel, built with its Series B funding. He says Mistral is investing heavily in infrastructure, with Series C and D clusters coming online soon to enable longer training, more ambitious post-training, and faster iteration — promising "large and rapid improvements" in the weeks and months ahead.
Related event: Mistral Trained ML4 on 3,800 Grace Blackwell GPUs in Europe(3 posts)→
More from Infra
- Microsoft's full Outlook outage PIR confirms AI tooling wasn't to blame — aronchick · 2026-11-22
- User reports 400+ tok/s with Flare inference engine — capetorch · 2026-11-22
- Huawei Ascend-950DT quoted at 3x 950PR price; yield and SuperPod interconnect may explain the gap — teortaxesTex · 2026-11-22
- AlphaChip and three TPU generations: a mine that digs itself is a throne — krishnan · 2026-11-22
- SK Hynix reportedly tops 70% yield on HBM4, starting mass production in Q3 — Beth_Kindig · 2026-11-22
- Mflux Runs MOSS Image Gen on 8GB Macs: 82s for 768x768 With 4-Step Distill LoRA — bio_risk · 2026-11-22