BobHuang
Beijing, China
Summary
👤
SeniorBob Huang is a compiler developer with eight years of experience specializing in ML and accelerator toolchains, currently adapting and optimizing the Triton compiler for next‑generation chip architectures. He has deep hands‑on expertise across Triton, MLIR, LLVM, Polygeist, PoCL and PyTorch integration, and has implemented an OpenCL (triton-spirv) backend for Triton while contributing to its codebase. Bob’s work spans operator development, performance tuning and inference optimization, and he has reverse‑engineered a many‑core GPGPU to extract and validate its instruction set—giving him rare visibility into both hardware microarchitecture and compiler lowering. He’s led practical porting efforts (including C model validation and QEMU‑based OpenCL simulation) to enable support for many‑core RISC‑V and proprietary chips, and has experience taking models from ONNX down to HLS-ready dialects. Based in Shanghai, he blends research-grade compiler transformations with production-focused integration and performance measurement.
8 years of coding experience
Github Skills (35)
inference10
large-language-models10
infer10
sequence-modeling10
bert9
question-answering9
natural-language-processing9
sentiment-analysis9
triton9
compression9
compiler9
embedding8
ptx8
duolingo8
information-extraction8
Programming languages (6)
Github contributions (5)
toyaix/tritonllm
Aug 2025 - Apr 2026
LLM Inference via Triton (Flexible & Modular): Focused on Kernel Optimization using CUBIN binaries, Starting from gpt-oss Model
Contributions:46 PRs, 104 pushes, 51 branches in 8 months
gptkernelllm-inferencetritoninfer
sBobHuang/HomeworkManage
Oct 2020 - Mar 2023
Contributions:100 pushes, 4 branches in 2 years 5 months
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.