Huawei Ascend Production Ramp: Die Banks, TSMC Continued Production, HBM is The Bottleneck

HBM, not logic, is the ceiling on China's accelerator output; that constraint shapes how much non-Western compute exists and where open-weight models can actually be trained and served.

• inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
• open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.

• articleAnother Giant Leap: The Rubin CPX Specialized Accelerator & Rack
• articleScaling the Memory Wall: The Rise and Roadmap of HBM
• articleH100 vs GB200 NVL72 Training Benchmarks – Power, TCO, and Reliability Analysis, Software Improvement Over Time

• postChina's Nvidia ban is a statement about its own siliconAndrew Ng
• podcastDylan Patel — Deep dive on the 3 big bottlenecks to scaling AI computeDwarkesh Patel
• postThe momentum case for China overtaking the US in AIAndrew Ng

Checking sign-in…

Loading comments…