Yizhu Technology Raises Over 1 Billion Yuan for Compute-in-Memory AI Chips

According to the report, rising AI compute demand has drawn capital-market and industry attention to new architectures such as general-purpose compute-in-memory, while Yizhu continues to advance technology development and product industrialization.

Yizhu began formal operations in 2022. It focuses on general-purpose compute-in-memory R&D and follows a route described as general-purpose compute-in-memory plus super-heterogeneous, with continuing work around 3D DRAM compute-in-memory and related directions.

Its products are described as offering high throughput, cost efficiency and low power consumption. They are compatible with mainstream ecosystems including CUDA and PyTorch, and are intended to provide a new compute solution for data centers and AI cloud computing with easier software use and lower token costs.

The financing will provide further support for Yizhu's subsequent R&D investment, product iteration and industrialization. As technical development and product deployment continue, the report said the company's general-purpose compute-in-memory route and related products will undergo fuller market validation. With AI compute demand growing, continued innovation in chip architecture, memory technology, software ecosystems and product applications is an important factor in moving new compute technologies toward industrial use.