Boyu Hao
Papers
1
Total Citations
5
H-Index
1
About
Boyu Hao is a leading researcher at the intersection of computer architecture, machine learning acceleration, and robotics. His work focuses on designing efficient, heterogeneous system-on-chip (SoC) solutions that bridge the gap between memory bandwidth and computational throughput for emerging AI workloads. Hao’s major contribution, the NeCTAr (Near-Cache Transformer Accelerator), is a groundbreaking 16nm RISC-V SoC that integrates both near-core and near-memory accelerators to handle sparse and dense machine learning kernels. This prototype, operating at 400MHz and 0.85V, achieves an impressive 109 GOPS for matrix-vector multiplications, demonstrating a scalable approach for language model inference and robotics. His related work, RASoC, further showcases his expertise in class-specific SoC design. With over 5 citations on his most recent publication, Hao’s research is gaining traction for its practical, silicon-proven solutions to real-time AI challenges. His achievements highlight a rare ability to translate theoretical advances into tangible hardware, making him a key figure in the push toward energy-efficient, high-performance edge computing.
Research Focus
Key Achievements
Top Papers
- 1