PhD ML Systems Engineer Intern: AI Hardware & Acceleration

AMD; San Jose (CA); Hybrid; USD 28,000 - 45,000; Full time

Job summary

AMD is seeking a highly motivated ML System Engineering co-op in San Jose/Santa Clara. You will own a project to accelerate LLM/VLM workloads on Ryzen AI NPU, study time/memory bottlenecks, and propose optimizations.

You will learn to write and tune AIE kernels with IRON to optimize GEMM, attention, and related operators. You will be paired with a mentor and will present your findings to the wider team, documenting lessons to guide future work.

Qualifications

• PhD in Computer Science, Computer Engineering, or Electrical Engineering.
• Strong knowledge of LLM and VLM architectures including attention and KV cache.
• Experience with compilers, graph lowering, operator fusion, scheduling, tiling, and memory planning; MLIR is a plus.
• Proficient in Python and C++, with ability to profile code to distinguish compute-bound vs memory-bound behavior.

Responsibilities

• Own a project to optimize an LLM or VLM on the Ryzen AI NPU; analyze bottlenecks in time and memory.
• Write and tune AIE kernels with IRON to optimize operators dominating runtime.
• Present findings to the engineering team and document learnings for future work.

Education

PhD in Computer Science, Computer Engineering, or Electrical Engineering

Tools

IRON

AMD AI Engine (AIE)

Job description

AMD is seeking a highly motivated ML System Engineering co-op in San Jose/Santa Clara. You will own a project to accelerate LLM/VLM workloads on Ryzen AI NPU, study time/memory bottlenecks, and propose optimizations.

You will learn to write and tune AIE kernels with IRON to optimize GEMM, attention, and related operators. You will be paired with a mentor and will present your findings to the wider team, documenting lessons to guide future work.