d-Matrix GA's Corsair rackscale, demos first public Attention-FFN disaggregation at 845 tok/s
ai · x · 2026-10-16
d-Matrix announced at OCP that Corsair rackscale is now generally available — air-cooled, industry-standard, deployable in existing data centers, built alongside Supermicro. It also introduced Squadron, multi-rack clusters scaling to 1,000+ Corsair XPUs, and ran the industry's first public demo of Attention-FFN Disaggregation: one NVIDIA H100 + 8 Corsair XPUs hitting 845 tokens/sec/user. The company bets on memory-centric inference compute.
More from Infra
- Muse-Spark open-weights drop: 816B-A42B MoE model gets 359GB MLX 3-bit build — ns123abc · 2026-10-16
- vLLM ships Day-0 LoRA RL for Muse Spark Open, lifting AIME from 0.56 to 0.85 — vllm_project · 2026-10-16
- DSpark draft model for Muse Spark Open proposes 7 tokens per pass with 23-42% acceptance — vllm_project · 2026-10-16
- Cloudflare Stream's WebRTC ultra-low-latency live streaming is now Generally Available — ritakozlov · 2026-10-16
- YC-backed Overshoot launches inference optimization, claims $5M saved for AI teams — ycombinator · 2026-10-16
- Meta's 816B open-weight Muse Spark Open hits Groq preview at 350+ tokens/sec — GroqLLC · 2026-10-16