Alibaba DAMO Academy Launches "Alibaba Star" Program with 15 AI and Chip Research Topics
Alibaba DAMO Academy has launched its "Alibaba Star" recruitment program targeting 2027 graduates, offering 15 advanced research topics. These topics span next-generation AI chips, RISC-V CPU architectures for AI agents, and multimodal medical AI models. This initiative highlights Alibaba's strategic focus on hardware-software co-design and specialized architectures like RISC-V to power the next generation of agentic AI. By targeting top-tier academic talent early, Alibaba aims to secure a competitive edge in the rapidly evolving AI chip and AGI landscape. The chip research tracks emphasize building custom compilers, optimizing inference systems for 100-billion-parameter models, and exploring RISC-V microarchitectures tailored for AI agent workloads. The AI track focuses heavily on practical applications, particularly multimodal AI agents for medical diagnostics and decision-making systems for heavy industries.
## BACKGROUND
An AI graph compiler translates high-level neural network computational graphs into optimized machine code for target hardware, bypassing traditional manual optimization. Hardware-software co-design is a methodology where hardware architecture and software stacks are developed simultaneously to maximize energy efficiency and performance, which is increasingly critical as AI workloads outpace general-purpose CPU capabilities.