A leading technology company is seeking a Principal Software Development Engineer to specialize in PyTorch training frameworks. This role involves optimizing AI training performance on AMD Instinct™ accelerators. The ideal candidate has significant experience in distributed training, deep knowledge of PyTorch internals, and can lead complex projects. Strong programming skills in Python and C++ are required. This hybrid position is based in San Jose, Seattle, or Austin, with a focus on innovation and technical leadership.
Qualifications
Experience with Autograd, optimizers, and compilation paths in PyTorch.
Ability to lead complex systems performance investigations.
Strong foundation in multi-node debugging and optimization.
Responsibilities
Act as a technical authority for PyTorch training at AMD.
Drive optimization of PyTorch training workloads.
Partner with teams to resolve performance and correctness issues.
Skills
Deep experience with PyTorch internals
Strong distributed training expertise
Proven track record in performance engineering
Strong programming skills in Python and C++
Clear technical communication skills
Education
Bachelor’s or Master’s degree in Computer Science, Computer Engineering, or Electrical Engineering
Tools
PyTorch
CUDA/HIP
Job description
A leading technology company is seeking a Principal Software Development Engineer to specialize in PyTorch training frameworks. This role involves optimizing AI training performance on AMD Instinct™ accelerators. The ideal candidate has significant experience in distributed training, deep knowledge of PyTorch internals, and can lead complex projects. Strong programming skills in Python and C++ are required. This hybrid position is based in San Jose, Seattle, or Austin, with a focus on innovation and technical leadership.