Justin Stoecker
Principal Software Engineer at Microsoft
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolJustin Stoecker is a Principal Software Engineer with 15 years of experience building low-level, hardware-accelerated ML platforms for GPUs and NPUs, currently focused on MLIR-based compilers for machine learning, real-time graphics, and gaming. At Microsoft he has advanced DirectML and the ONNX Runtime DirectML execution provider, adding performance-critical features like native FP16 DFT kernels and STFT support that bridge shader development and ML inference. His work spans production-grade GPU acceleration, ONNX model integration, and tooling improvements for cross-vendor DirectX 12 hardware. Holding graduate training from the University of Miami, he combines deep academic roots with pragmatic systems engineering and a knack for surfacing subtle cross-platform issues (e.g., WSL path handling) that improve developer workflows.
15 years of coding experience
10 years of employment as a software developer
Ph.D. Computer Science, Ph.D. Computer Science at University of Miami
Stackoverflow
98reputation
6kreached
4answers
1question
Github Skills (16)
mle10
c-language10
cprogramming-language10
gpu-acceleration10
onnx10
directml10
directx10
ml10
shader9
machine-learning9
direct3d6
slimdx6
pixel6
rendering6
direct2d6
Programming languages (6)
Github contributions (5)
microsoft/DirectML
Jun 2020 - Nov 2022
DirectML is a high-performance, hardware-accelerated DirectX 12 library for machine learning. DirectML provides GPU acceleration for common machine learning tasks across a broad range of supported hardware and drivers, including all DirectX 12-capable GPUs from vendors such as AMD, Intel, NVIDIA, and Qualcomm.
Role in this project:
ML Engineer
ML EngineerContributions:2 releases, 159 reviews, 45 commits in 2 years 5 months
Contributions summary:Justin contributed to the DirectML project by addressing issues and enhancing the tools related to machine learning acceleration on DirectX 12 capable GPUs. Their commits include fixes for path separators in WSL environments, improvements to the YOLO sample, and updates to the DxDispatch tool, particularly involving ONNX model handling. The user demonstrated an understanding of DirectML libraries and ONNX model integration.
amdaccelerationdirectmlgpu-accelerationnvidia
microsoft/onnxruntime
Sep 2020 - Oct 2022
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
Role in this project:ML Engineer
ML EngineerContributions:47 reviews, 26 commits, 26 PRs in 2 years
Contributions summary:Justin contributed significantly to the DirectML execution provider within the ONNX Runtime project. Their work included integrating the STFT (Short-Time Fourier Transform) operator by implementing it as a custom op, which involved implementing a DML Mul/Identity + DFT kernel. They also added native FP16 shader support to the DFT kernel, along with addressing SDL warnings and enabling graph fusion. The user's contributions demonstrate expertise in extending the DML EP and optimizing performance for GPU acceleration.
runtimetrainingtensorflowai-frameworkaccelerator
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.