"The CUDA moat is gone": Japanese neocloud ai& deploys Tenstorrent at scale
DavidBennett__ · x · 2026-09-13
David Bennett, CEO of Japanese neocloud ai&, sat down with analyst Ian Cutress. Key points:
- Bennett claims "the CUDA moat is gone";
- A chunk of ai&'s infrastructure runs on Tenstorrent hardware, one of the largest deployments outside the US;
- The company bets on heterogeneous compute, sovereign AI and surging inference demand;
- As AI becomes agentic, "one person no longer means one token" — token consumption is scaling fast.
Related event: Japanese neocloud ai& claims CUDA moat is gone, bets on Tenstorrent(2 posts)→
More from Infra
- DeepMind Chief Strategist: AI Infrastructure Spending Is a Bet on Recursive Self-Improvement — rohanpaul_ai · 2026-09-13
- AI valuations can't all be right: memory at 3-5x PE vs premium infrastructure, says Gavin Baker — rohanpaul_ai · 2026-09-13
- Draw Things launches Local Code beta: local coding agents on Mac at ~980 tok/s prefill — liuliu · 2026-09-13
- Macrocosmos launches IOTA for liquid training on scattered, disaggregated compute — markjeffrey · 2026-09-13
- Dev launches Lorivo: one GPU server serves many LoRA adapters via vLLM — TheOneWhoWil · 2026-09-13
- Patched vLLM+FlashInfer Pushes Gemma 4 31B to 150 tok/s on a Single B300, Beating SGLang — abhijithneil · 2026-09-13