Bittensor weekly: lium hosts $30M in GPUs, affine ships ~10x cheaper video model
markjeffrey · x · 2026-10-24
Weekly recap of the Bittensor network:
- @liumio now hosts $30M in GPUs (91% utilization), including 225 B300s, at a $21M run rate
- @affineio dropped a video model that is 10x cheaper than the base model with virtually no quality loss after 3 days of optimization
- @engyai hit a new ATH, processing 370B tokens this week
- @saygm had its highest-traffic week, crossing $1M in total volume since July
- @cacheonai added Meta's Muse Spark Open Kernel challenge to its mining competition
- Ingot released Ingot-I, a Qwen 27B FP8 quantization that beats Alibaba's own quantization, built in just 1 week
Related event: Bittensor Weekly: Lium Hosts $30M of GPUs at 91% Utilization(3 posts)→
More from Infra
- NVIDIA makes 8-node DGX Spark clusters official: 97.4% scaling efficiency on 900M training — TheZachMueller · 2026-10-24
- One NVIDIA Vera Rubin rack would set you back at least ~$7M — LinusEkenstam · 2026-10-24
- DSF4.1 hits 280 tok/s on a single RTX 6000 — community chases 500 tok/s setups — capetorch · 2026-10-24
- PS5 Pro sells out in Toronto as memory shortage hits consumer hardware — Rasmic · 2026-10-24
- Open-source kapitsa engine boosts Mac local agent throughput 23% with tree speculative decoding — DotaMate · 2026-10-24
- Nvidia compute unlock boosts RTX 3080 20GB prefill by up to 29%, decode unchanged — No-Refrigerator-1672 · 2026-10-24