Nathaniel Pinckney

I'm Nate Pinckney. I received my Ph.D. in Electrical Engineering from the University of Michigan in 2015 and am currently a Senior Research Scientist at NVIDIA in Austin, Texas. My work spans deep learning accelerators, high-level synthesis methodologies, low-power VLSI design, and cryptographic hardware, with over 40 publications in these areas. I have full flow experience from architecture down to physical implementation and circuits.

I am a co-author of the industry-adopted Connections and MatchLib SystemC HLS libraries. More recently, my research has focused on evaluating large language models and AI agents for hardware design, including co-creating the VerilogEval and Comprehensive Verilog Design Problems (CVDP) benchmarks.

I lead collaborative initiatives across industry and academia, serving as Chair of the Si2 LLM Benchmarking Coalition, Sponsorship/Industry Relations Chair for the International Conference on LLM-Aided Design (ICLAD), and as a technical subcommittee member for both DAC and ISSCC.

I'm an Eagle Scout. In my spare time, my wife and I enjoy playing board games, playtesting prototypes, and attending many board game conventions across the US.