Machine Learning Stack Integration Engineer

The Role:

AMD is looking for people who will be working with Solution Validation & Debug within the Machine Learning Software Engineering group. As a team member responsibility will include working closely on the debug and triage of Machine learning and High-Performance Computing related issues that contribute to the Solution Validation of ROCm Stack.

The Person:

The ideal candidate will bring broad experience in dealing with sophisticated software level issues related to Machine Learning and High-Performance Computing. They are self-motivated, have outstanding problem-solving skills, will thrive in a fast-paced environment, and have a validated ability to collaborate within and across diverse teams.

Key Responsibilities:

• Work on ROCm stack packaging solution for individual and enterprise level deployment on distributed cloud infrastructure
• Debug Machine Learning/ High Performance Computing related issues on Radeon Open Compute Stack (ROCm)
• Develop test contents for sophisticated Machine Learning algorithms on distributed nodes
• Port High Performance computing application on ROCm
• Reproduce field defects and develop appropriate tests to prevent future issues.
• Design, develop and deploy testing tools and automation libraries vital to perform testing.
• Be responsible for the adoption of tooling and industry standard methodologies by means of advocacy and outreach to help our development communities’ level up.

Preferred Experience:

• Languages: Python, C, C++, Linux Shell scripting.
• Frameworks/Libraries: TensorFlow, PyTorch, ONNXRT
• Tools: Prior experience with Linux, Docker, LLVM compilers, GNU make /CMAKE, Jenkins, Git/Gerrit
• Understanding of High-Performance Computing application, Machine learning and GPU Programming, MPI Parallel Programming

Academic Credentials:

Bachelor's Degree in Computer Science or related quantitative field. An advanced degree or equivalent practical work experience is a plus.