GPU Software Engineer — HPC & ML Libraries Lead

Socket.dev; California (MO); Hybrid; USD 180,000 - 230,000; Full time

Job summary

AMD in San Jose, CA, invites a Senior Member of Technical Staff to advance ROCm GPU libraries strategy and execution. You will work with engineers and applied mathematicians on high-performance computing and machine learning workloads, shaping software that spans architecture to implementation.

You will collaborate across teams, design and deliver critical library capabilities, mentor staff, and drive performance-oriented software that aligns with roadmap commitments in a hybrid work setup.

Qualifications

• Hands-on C++ with CUDA/HIP/OpenMP/MPI/OpenCL for GPUs and CPUs.
• Familiarity with Python and AI/ML frameworks.
• Experience with open-source software development and agile methods.

Responsibilities

• Design, implement, validate, and deliver key software library capabilities across ROCm ecosystem.
• Guide technical execution and contribute to architecture and documentation efforts.
• Collaborate with senior staff to develop optimal designs considering performance and roadmap commitments.
• Mentor junior staff and promote best practices in agile environments.

Skills

C++

CUDA

HIP

OpenMP

MPI

OpenCL

Python

AI/ML frameworks

Education

Master's degree or higher

Tools

Jenkins

Docker

Kubernetes

Job description

AMD in San Jose, CA, invites a Senior Member of Technical Staff to advance ROCm GPU libraries strategy and execution. You will work with engineers and applied mathematicians on high-performance computing and machine learning workloads, shaping software that spans architecture to implementation.

You will collaborate across teams, design and deliver critical library capabilities, mentor staff, and drive performance-oriented software that aligns with roadmap commitments in a hybrid work setup.