Vamber Yang
Papers
1
Total Citations
5
H-Index
1
About
Vamber Yang is a leading architect of next-generation silicon for artificial intelligence, whose work bridges the gap between processor design and real-world machine learning deployment. Yang’s primary research areas are energy-efficient computer architecture, near-memory computing, and domain-specific accelerators for AI and robotics. Their most notable contribution is the NeCTAr (Near-Cache Transformer Accelerator), a 16nm heterogeneous multicore RISC-V system-on-chip (SoC) that integrates both near-core and near-memory accelerators to efficiently handle sparse and dense machine learning kernels. This prototype chip, operating at 400MHz and 0.85V, achieves a remarkable 109 GOPS for matrix-vector multiplications, demonstrating a powerful approach to reducing data movement—the primary bottleneck in modern AI inference. Co-developed with the RASoC robotics SoC, this work showcases a dual-path strategy for language model inference and autonomous systems. With 5 citations already in its first year, this 2024 paper signals a high-impact trajectory. Yang’s work is essential reading for students and researchers interested in the future of edge AI, where hardware and algorithms must co-evolve to deliver intelligence efficiently, from cloud servers to autonomous robots.
Research Focus
Key Achievements
Top Papers
- 1