Report: DeepSeek training a 2T-param model with 8T planned, Huawei chips due late 2026
Hesamation · x · 2026-09-21
According to a report, DeepSeek is training a 2T-parameter model and has an 8T model planned. For context, Kimi K3 is 2.8T and DeepSeek V4 Pro is 1.6T.
The company also reportedly expects new Huawei training chips in Q4 2026 or Q1 2027, with the CEO telling investors that moving to domestic-chip training “has to work.” Note: unverified rumor.
More from Infra
- Photoroom reclaims ~1,000 idle GPU hours a week with SkyPilot across Slurm and Kubernetes — skypilot_org · 2026-11-12
- Cloudflare Talk: Swapping Some Agent LLM Calls for Fast Models Like Jev Gives Big Speedups — michellechen · 2026-11-12
- Cloudflare Trains Its Own Clef and Clairic Models to Complete Its Agent Stack — michellechen · 2026-11-12
- AWS open-sources 23 production-ready GenAI assets, cutting build time from months to weeks — AWS ML Blog · 2026-11-12
- Sixth Street CEO: ~$200B of Chips Sitting in Warehouses as Data Centers Lag Behind — SumitGup · 2026-11-12
- AWS CEO Matt Garman: Space Data Centers Are 10-20 Years Away, Likely Not Economically Feasible — rohanpaul_ai · 2026-11-12