Songchen Tan
Research Assistant at Massachusetts Institute of Technology
Cambridge, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolSongchen Tan is a doctoral researcher at MIT CCSE and a research assistant in CSAIL’s Julia Labs, bringing eight years of experience at the intersection of mathematical theory, high-performance computing, and machine learning. They design scalable automatic differentiation algorithms and integrate AD at the LLVM IR level via the Enzyme project, enabling cross-language differentiation for languages like Julia, C++, and Fortran. Songchen has a strong HPC pedigree—contributing to electronic-structure codebases and optimizing distributed simulation and BLAS/LAPACK derivative kernels—while also shipping production-grade C++ inference backends and compiler optimizations that yielded measurable speedups. Their work blends rigorous theory with pragmatic engineering: synthesizing linear-algebra-aware derivative kernels, improving layer fusion heuristics for deep-learning compilers, and dramatically reducing memory issues in cloud-native deployments. Based in Cambridge with a PhD focus in Computational Science and Engineering, they combine open-source contributions and deep systems work to advance both research and deployable tools.
8 years of coding experience
1 year of employment as a software developer
Doctor of Philosophy - PhD, Computational Science and Engineering, Doctor of Philosophy - PhD, Computational Science and Engineering at Massachusetts Institute of TechnologyHigh School Diploma, High School Diploma at The High School Affiliated to Renmin University of ChinaBachelor of Science - BS, Chemistry & Physics, Bachelor of Science - BS, Chemistry & Physics at Peking University
Github Skills (13)
ec10
fortran10
struct10
c-language10
computational-physics10
cmake10
cprogramming-language10
linear-algebra9
parallel-computing9
data-serialization8
openmp8
mpi8
serialization8
Programming languages (14)
Github contributions (5)
abacusmodeling/abacus-develop
May 2021 - Jul 2021
An electronic structure package based on either plane wave basis or numerical atomic orbitals.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:37 commits in 1 month
Contributions summary:Songchen's commits primarily focus on modifying source code related to the ABACUS electronic structure package. These modifications include adding and updating CMake files for build configuration and integrating changes from the master branch. Additionally, the user has modified core computational modules, such as those related to molecular dynamics, and local potential calculations, indicating involvement in implementing or improving core functionality within the package. Further commits include changes related to the serialization of data.
chemical-engineeringelectronic-structureplane-waveorbitalspackage-based
JuliaDiff/TaylorDiff.jl
Nov 2022 - Feb 2023
Taylor-mode automatic differentiation for higher-order derivatives
Contributions:60 commits, 11 PRs, 54 pushes in 3 months
automatic-differentiationdifferentiationpdetaylor-seriesautodiff
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.