About

I am a Ph.D. Candidate in Computer Science at Harvard University’s School of Engineering and Applied Sciences. I work in the Edge Computing Lab advised by Prof. Vijay Janapa Reddi. My research focuses on two themes: Bespoke AI and AI for Computer Architecture.

During my Ph.D., I have spent summers interning with Google’s Open-Source Silicon Team, Pragmatic Semiconductor’s R&D Group, and NVIDIA’s Architecture Research Group to advance my research. Prior to Harvard, I received my B.S. in Computer Engineering from Columbia University.

Research

Bespoke AI

Hardware-software co-design for ultra-low power systems to enable machine learning on Extreme Edge and emerging-technology platforms.

Key Projects & Contributions
  • Bendable Non-Silicon RISC-V Microprocessor (Nature 2024)
  • Lifetime-Aware Design for Item-Level Intelligence at the Extreme Edge (ASPLOS 2026; Best Paper Award)
  • 333-eDRAM-3T Embedded DRAM (DAC 2025)
  • CFU Playground (ISPASS 2023)
  • Is TinyML Sustainable? (CACM 2023)

AI for Computer Architecture

Building AI agents and ML-based methodologies that accelerate hardware architecture exploration, design, and evaluation.

Key Projects & Contributions
  • QuArch v1.0 (ICML 2026)
  • QuArch v0.1 (CAL 2025; HPCA 2026 Best of CAL)
  • ArchGym (ISCA 2023)
  • PerfCompass

News

  • Apr 2026 “QuArch: A Benchmark for Evaluating LLM Reasoning in Computer Architecture” (QuArch v1.0) accepted to ICML 2026.
  • Apr 2026 Named a 2026 ML and Systems Rising Star by MLCommons.
  • Mar 2026 Received the Best Paper Award at ASPLOS 2026 for “Lifetime-Aware Design for Item-Level Intelligence at the Extreme Edge.”
  • Jan 2026 “QuArch: A Question-Answering Dataset for AI Agents in Computer Architecture” (QuArch v0.1) received the Best of CAL Award at HPCA 2026.
  • Jan 2026 Gave an invited talk on Architecture 2.0 at the DISCC Workshop, HPCA 2026.
  • Dec 2025 Selected as a recipient of the NVIDIA Graduate Fellowship for the 2026–2027 academic year.
  • Aug 2025 Completed second Ph.D. Research Internship in the Architecture Research Group at NVIDIA.
  • May 2025 Winners announced at MLArchSys, ISCA 2025 for the first QuArch/ArchReasoning Competition.
  • Sep 2024 “Bendable Non-Silicon RISC-V Microprocessor” published in Nature 2024.