Hyoukjun Kwon
Papers
2
Total Citations
470
H-Index
2
About
Hyoukjun Kwon is a leading researcher at the intersection of computer architecture and deep learning, renowned for pioneering efficient hardware accelerators. His most impactful contribution is **SIGMA**, a sparse and irregular GEMM accelerator with flexible interconnects for DNN training, which has garnered **462 citations** for its transformative approach to handling the non-dense computations central to modern neural networks. This work directly addresses the critical challenge of efficiently supporting sparsity and irregular data patterns in deep learning hardware, enabling faster and more energy-efficient training. Beyond hardware acceleration, Kwon has also contributed to hardware security, developing automated methods to prove flow security in sequential logic through synthesizable relational invariants. His research spans the full stack from algorithms to circuits, making him a key figure in the push toward specialized, high-performance computing systems. With a focus on practical, high-impact solutions, Kwon’s work continues to shape how next-generation AI workloads are executed, bridging the gap between algorithmic innovation and efficient hardware implementation.
Research Focus
Key Achievements
Top Papers
- 1
- 2