Lead GPU Performance Engineer for AI Training and Finetuning

AMD; San Jose (CA); Hybrid; USD 120,000 - 160,000; Full time

Job summary

A leading semiconductor company is seeking a GPU Performance Engineer to enhance training pipelines for AMD GPUs. The ideal candidate will have expertise in deep learning frameworks like PyTorch and experience optimizing multi-GPU training. Strong skills in Python and C++ are required. The position is hybrid, and applicants from various locations in the US will be considered. This role offers the chance to lead significant performance improvements in AI workloads.

Qualifications

• Experience with GPU performance for deep learning applications.
• Hands-on experience with SFT, LoRA, and RL-based training solutions.
• Strong background in distributed systems and communication libraries.

Responsibilities

• Lead performance for finetuning and reinforcement learning training solutions.
• Optimize multi-GPU training and communication patterns.
• Ship reproducible pipelines and document results.

Skills

GPU performance engineering

Deep learning frameworks

Distributed systems

Python

C++

Education

B.S./M.S./Ph.D. in Computer Science, Computer Engineering, or Electrical Engineering

Tools

ROCm

Triton

PyTorch

Job description

A leading semiconductor company is seeking a GPU Performance Engineer to enhance training pipelines for AMD GPUs. The ideal candidate will have expertise in deep learning frameworks like PyTorch and experience optimizing multi-GPU training. Strong skills in Python and C++ are required. The position is hybrid, and applicants from various locations in the US will be considered. This role offers the chance to lead significant performance improvements in AI workloads.