Yilong Guo
Compiler Engineer at Intel Corporation
Minhang District, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolYilong Guo is a compiler engineer with 11 years of experience, currently working at Intel on DPC++ and OpenCL technologies. Trained in microelectronics at Tsinghua University (BS/MS), he focuses on low-level translation and tooling—evidenced by contributions to the Khronos SPIRV-LLVM Translator where he implemented OpenCL 1.2/2.0 interfaces and fixed tricky debug-info and builtin translation bugs. He also brings practical ML deployment expertise, optimizing Intel IPEX paths for the popular SD.Next generative image project to accelerate inference on Intel hardware. Known for a “build what's missing” mindset, he blends compiler internals with pragmatic performance engineering to close gaps between IR, runtime, and hardware. Colleagues rely on him for robust cross-stack fixes that improve both correctness and throughput in production toolchains.
11 years of coding experience
Master's degree, Microelectronics, Master's degree, Microelectronics at Tsinghua University
Github Skills (24)
spirv10
opencl10
pytorch10
c-language10
intel10
python10
llvm10
stable-diffusion10
optimisation10
ai10
ai-image10
generative-art10
cprogramming-language10
optimization10
diffusers9
Programming languages (11)
Github contributions (5)
KhronosGroup/SPIRV-LLVM-Translator
Feb 2021 - Sep 2021
A tool and a library for bi-directional translation between SPIR-V and LLVM IR
Role in this project:
Back-end Developer
Back-end DeveloperContributions:15 reviews, 21 commits, 9 PRs in 7 months
Contributions summary:Yilong's contributions center around enhancing the SPIRV-LLVM translator, focusing on the correct implementation of SPIR-V to OpenCL translation. They introduced new interfaces and classes for OpenCL 1.2 and 2.0, and fixed an issue related to debug information handling during translation. Furthermore, the user addressed bugs in SPV-IR builtin variable/call translation, and made improvements to the translation of work-item builtins to variables.
llvm-irtranslationbi-directionalspir-vllvm
vladmandic/sdnext
Jun 2023 - Nov 2023
SD.Next: All-in-one for AI generative image
Role in this project:ML Engineer
ML EngineerContributions:1 review, 6 PRs, 17 comments in 4 months
Contributions summary:Yilong primarily focused on integrating and optimizing Intel's IPEX (Intel Extension for PyTorch) backend within the SD.Next project to accelerate AI-based image generation. Their contributions include modifying code to support IPEX for SDE samplers, fixing batch normalization issues for Tiled VAE, and addressing ControlNet related configurations for depth processing. Additionally, the user implemented and fixed native windows setups. The impact of the user's work is to optimize model performance for Intel hardware.
pytorchgenerative-artstablediffusiondiffusiontext2image
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.