Software Development Engineer - GPU Libraries

Advanced Micro Devices, Inc • San Jose, CA

3d

About The Position

AMD is seeking a Software Engineer to help develop open-source AI software for GPU accelerators used in supercomputers and data centers. The role contributes to technology focused on GPU performance and efficiency and supports AI workloads across data center GPUs. The engineer will work with open-source projects and collaborate with partners and communities. The role also involves developing internal AI utilities, automated agents, and developer workflows, using large language models to assist with code generation, custom-kernel optimization, and debugging GPU performance bottlenecks. A bachelor’s or master’s degree in a relevant engineering or computer science field, or equivalent, is listed under academic credentials. The role is not eligible for visa sponsorship.

Requirements

• Proficiency in C++, Python, or similar programming languages.
• Strong foundational knowledge of AI training, inference methodologies, and full-stack AI tool development and lifecycle integration.
• Familiarity with GPU programming, including CUDA, HIP, or OpenCL, and performance optimization techniques.
• Bachelor’s or master’s degree in Computer Science, Computer Engineering, Electrical Engineering, or equivalent.

Responsibilities

• Contribute to ROCm, Verl, Ray, Triton Inference Server, FlashInfer, physical simulation, and other emerging open-source projects.
• Collaborate with leading partners and open-source communities to enable AI workloads and improve performance on data center GPUs.
• Architect, build, and deploy custom internal AI utilities, automated agents, and developer workflows.
• Use advanced large language models to accelerate code generation, optimize custom kernels, and automate debugging of GPU performance bottlenecks.

What This Job Offers

• Job Type: Full-time