David (Seung Won) Min

About Me

I am a GPU Architect at NVIDIA. I received my Ph.D. in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign, advised by Prof. Wen-mei Hwu.

My research centers on accelerator-centric data communication, with a focus on improving data transfer efficiency in heterogeneous CPU-GPU and CPU-FPGA systems spanning hardware interconnects, operating systems, and applications. Prior to NVIDIA, I interned at IBM T.J. Watson Research Center, where I worked on high-performance CPU-FPGA communication over the Coherent Accelerator Processor Interface (CAPI).

Research Interests

• Accelerator-Centric Data Communication: improving data transfer efficiency across heterogeneous CPU-GPU and CPU-FPGA systems
• Graph Neural Network Training: GPU-centric data access for large-scale GNN workloads
• FPGA Systems: I/O cache coherency, hardware-software co-design, and FPGA-based prototyping

Talks

• Unified Tensor - Enabling GPU-Centric Data Access for Efficient Large Graph GNN Training
  Graph Neural Networks User Group Meeting, Sept. 2021 [video]

• PyTorch-Direct: Introducing Deep Learning Framework with GPU-Centric Data Access for Faster Large GNN Training
  NVIDIA GTC 2021 [video]

Publications

Powered by Jekyll and Minimal Light theme.