Yutaro Iiyama
Assistant Professor at ICEPP, The University of Tokyo
Tokyo, Japan
Summary
👤
Senior🎓
Top SchoolYutaro Iiyama is an Assistant Professor at The University of Tokyo with 12 years of experience in experimental particle physics, specializing in high-energy searches for novel particles with the CMS experiment at the LHC. He combines deep physics analysis expertise with hands-on software and DevOps skills, having contributed to Monte Carlo generator configurations and production workflows at CERN and beyond. His open-source work on hls4ml demonstrates practical FPGA-ML integration, including converter and template refactors that enable nonlinear model support and novel quantization schemes. Trained with a Ph.D. from Carnegie Mellon and seasoned by roles at MIT and CERN, he bridges cutting-edge detector physics and deployable ML/FPGA tooling. Notably, his background spans both theoretical generator tuning and low-level hardware-aware ML optimization, making him adept at translating complex simulations into production-ready systems.
12 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy (Ph.D.), Experimental Particle Physics, Doctor of Philosophy (Ph.D.), Experimental Particle Physics at Carnegie Mellon University
Github Skills (9)
neural-network10
fpga10
keras10
machine-learning10
vivado10
python10
bash9
pytorch8
onnx8
Programming languages (6)
Github contributions (5)
cms-sw/genproductions
May 2016 - Apr 2019
Generator fragments for MC production
Role in this project:
Backend & DevOps Engineer
Backend & DevOps EngineerContributions:31 commits, 10 PRs, 21 comments in 2 years 10 months
Contributions summary:Yutaro contributed to the generation of Monte Carlo production configurations, including modifications to MadGraph5_aMCatNLO scripts and patches. They updated the configuration files for the JHUGen generator and created new hadronizer configurations. Additionally, the user modified scripts related to gridpack generation and PDF set handling, demonstrating a focus on optimizing the production workflow and integrating different tools.
productionfragments
fastmachinelearning/hls4ml
Mar 2020 - Aug 2021
Machine learning on FPGAs using HLS
Role in this project:ML Engineer
ML EngineerContributions:6 reviews, 12 commits, 8 PRs in 1 year 5 months
Contributions summary:Yutaro primarily focused on enhancing the `hls4ml` library to support more complex machine learning models. They made several code modifications to the `keras_to_hls.py` converter, notably making input and output shapes explicit for nonlinear models. Additionally, the user reorganized and refactored Vivado templates, including updates for defining precision, handling input data, and implementing new quantization schemes, showcasing their expertise in FPGA-based machine learning model conversion.
pytorchpythonvivadofpgasonnx
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.