Browsing: semiconductors

Alibaba targets massive model training with a new accelerator architecture and an ambitious 20-gigawatt infrastructure roadmap.

The company reported 42% revenue growth as it shifts toward autonomous design platforms and royalty-based intellectual property models.

The Raptor accelerator utilizes direct-bonded logic and DRAM to overcome the memory wall for large-scale AI inference.

The Shanghai-based startup is pivoting to an open-weight strategy for its latest video model, aiming to disrupt the dominance of proprietary platforms in the Chinese AI market.

Meituan has released a massive sparse Mixture of Experts model trained on domestic Chinese ASIC clusters to reduce reliance on Western hardware.

OpenAI is moving to design its own specialized inference hardware to reduce reliance on third-party GPU suppliers. The project aims to optimize model performance through custom silicon architecture.