Yi Zhang - Software Engineer at Google

Yi Zhang

Software Engineer at Google

Canada

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Yi Zhang is a software engineer with a decade of experience specializing in JIT compilers, virtual machine runtimes, and ML compilers, currently contributing to Google’s ML compiler efforts. He spent eight years at IBM developing OpenJ9/OMR Java runtime features—owning critical optimizations like the inliner, interpreter abstractions, and runtime patching that improved throughput and debugging performance. Skilled in low-level performance investigation, live debugging and postmortem analysis on Windows and Linux, he pairs deep knowledge of CPU architecture with pragmatic compiler engineering. At Google he has bridged systems and ML by contributing backend work to high-profile projects such as Torch-MLIR and IREE, implementing end-to-end PyTorch op support and model conversion for real inference workloads. Notable for shipping cross-architecture backend support (x86, PowerPC, Z) and integrating torch-mlir dialects into compiler toolchains, he excels at taking research ideas into production-grade compiler passes. Based in Canada with an M.Sc. from McGill, he combines academic rigor with hands-on systems and ML tooling experience.

10 years of coding experience

9 years of employment as a software developer

Bachelor of Engineering (B.Eng.) Measurement and Control, Bachelor of Engineering (B.Eng.) Measurement and Control at Harbin Institute of Technology

Master of Science (M.Sc.) Computer Science, Master of Science (M.Sc.) Computer Science at McGill University

Github Skills (13)

compiler10

pytorch10

machine-learning10

compiler-compiler10

python10

mlr10

tensorflow9

back-end-development9

c-language8

mle8

ml8

cprogramming-language8

mlops5

Programming languages (8)

TypeScriptJavaC++ShellLLVMGoMLIRPython

Github contributions (5)

llvm/torch-mlir

Sep 2021 - May 2022

The Torch-MLIR project aims to provide first class support from the PyTorch ecosystem to the MLIR ecosystem.

Role in this project:

Back-end Developer & ML Engineer

Contributions:635 reviews, 43 commits, 139 PRs in 8 months

Contributions summary:Yi implemented end-to-end support for several PyTorch operations (e.g., `aten.cat`, `aten.gather`, `aten.bmm`), including the conversion of TorchScript models to MLIR. They focused on refining the type system, especially for scalar values and tensor operations. Furthermore, the user added support for the MiniLM-L6-H384-uncased-sst2 model and a softmax backward module, demonstrating their contributions to integrating and expanding the library's capabilities for machine learning model conversion and inference.

pytorchmlirtorchcompilerecosystem

iree-org/iree

Jul 2021 - Apr 2022

A retargetable MLIR-based machine learning compiler and runtime toolkit.

Role in this project:

Backend Developer

Contributions:32 reviews, 10 commits, 8 PRs in 9 months

Contributions summary:Yi primarily contributes to the IREE compiler, focusing on input conversion and optimization passes. Their work includes implementing a pattern to convert tensor casts to flow tensor operations, improving the compiler's ability to handle various tensor transformations. They also integrated the torch-mlir-dialects into the IREE build process, enabling the compilation of models from PyTorch. These contributions enhance the compiler's functionality and expand its support for different machine learning frameworks.

mlirspirvvulkantensorflowcompiler

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.

Request Free Trial