THE ROLE:

• We are looking for an experienced Machine Learning GPU driver performance engineer to optimize and develop ML GPU drivers and kernels with a focus on improving performance of DirectML content creation applications.
• This is an excellent opportunity to work on the cutting edge of Machine Learning and GPU architecture.

THE PERSON:

• You are passionate about machine learning and improving the performance of games and applications utilizing the GPU for machine learning acceleration.
• Has hands-on experience and understanding of GPU architecture
• Has good programming skills in HLSL, GLSL, OpenCL, CUDA or similar
• Has solid general software development skills in C++
• Able to communicate effectively and work optimally in a distributed team  
• Multiple office locations available

KEY RESPONSIBILITIES:

• Analyze shader algorithms in HLSL and other shading languages, analyze data from performance tools, identify issues and potential optimization opportunities, and develop algorithms/shaders to achieve performance improvement.
• Brainstorm with teams about potential solutions to optimize, experiment and refine optimizations and implement in drivers.
• Contribute to performance tool improvements.

PREFERRED EXPERIENCE:

• Good understanding of GPU architecture and SIMD execution
• Good understanding of machine learning runtimes, Microsoft Olive, ONNX, TensorFlow or PyTorch
• Hands-on experience in developing machine learning drivers on D3D, Vulkan or Metal APIs.
• Experience in HLSL, GLSL, OpenCL, CUDA or similar and strong C/C++ programming skills
• Working understanding of memory management, and operating systems fundamentals

• Excellent communication and collaboration skills to work across teams

#LI-HS3
#LI-DNI = Do Not Post on LinkedIn