Vijay Kandiah
Senior Compiler Developer at NVIDIA
Chicago, Illinois, United States
Summary
🤩
Rockstar🎓
Top SchoolVijay Kandiah is a Senior Compiler Developer at NVIDIA with seven years of experience specializing in HPC compilers at the intersection of computer architecture and energy-efficient parallel computing. He holds an MS and is pursuing a PhD in Computer Engineering from Northwestern University, where he spent five years as a research assistant working on parallel architecture and compiler co-design. Vijay’s internship work at NVIDIA produced high-impact optimizations—SPMD C++ extensions and an LLVM IR-to-IR pass that matched hand-written intrinsics and achieved up to 2.6x performance vs. auto-vectorization—and he now contributes to the LLVM project’s Flang front end, implementing FIR optimizer hooks and codegen improvements. Based in Chicago, he blends deep research experience with production compiler engineering, notably enabling flexible GPU kernel generation and aggressive alloca hoisting for tangible speedups in real-world HPC workloads.
7 years of coding experience
6 years of employment as a software developer
Doctor of Philosophy - PhD Computer Engineering, Doctor of Philosophy - PhD Computer Engineering at Northwestern UniversityBachelor of Science (B.S.) Electrical Electronics and Communications Engineering, Bachelor of Science (B.S.) Electrical Electronics and Communications Engineering at Anna University Chennai
English, Tamil, French
Github Skills (9)
compiler-design10
flang10
optimization10
llvm10
openacc9
cprogramming-language9
c-language9
fortran9
mlr8
Programming languages (5)
Github contributions (5)
llvm/llvm-project
Jan 2024 - Nov 2024
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:8 reviews, 16 PRs, 6 pushes in 9 months
Contributions summary:Vijay primarily contributes to the Flang front-end compiler within the LLVM project. Their work includes adding extension point callbacks to the FIR optimizer pipeline, enabling the registration of inliner passes. They modified the codegen to hoist constant-sized allocas for improved optimization and resolved issues with alloca reordering. The user also added a custom builder for the acc::ParallelOp, allowing more flexible operand specification.
compilerstechnologiesclangsubmittoolchain
VijayKandiah/LLVM_CAT
Oct 2019 - Nov 2019
Contributions:14 commits, 13 pushes, 1 branch in 19 days
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.