Alibaba Group Holding’s chip design unit T-Head announces it is open-sourcing SAIL, the full software stack used with its Zhenwu series of AI chips. The announcement comes at the World AI Conference (WAIC) in Shanghai. According to the report, SAIL is presented as the foundational software architecture for the Zhenwu chips, and T-Head says it is making the complete technical components of the stack available for developers. The stated goal of the release is to make it easier for programmers to work with T-Head’s AI hardware and reduce the effort required to move away from Nvidia’s CUDA-based ecosystem. Both outlets describe the initiative as a bid to address “lock-in” concerns for developers who are already using CUDA and to streamline development workflows around alternative AI chip platforms. The coverage characterizes the move as part of T-Head’s broader efforts to broaden adoption of its AI chips by improving software accessibility and interoperability, though details on how widely SAIL is supported beyond mainstream developer environments are not fully specified in the excerpts.