Software Engineer, Systems ML - Frameworks/Compilers/Kernels (PhD)

Meta is seeking a Software Engineer for their PyTorch Accelerator Enablement (AE) team, focusing on optimizing GenAI model inference on GPUs. This role is part of the industry-leading PyTorch AI framework organization, where you'll work on developing and implementing various optimization techniques at the algorithm, kernel/compiler, and runtime levels.

The position requires a PhD in Computer Science or related field, with strong programming skills in Python and C/C++. You'll be at the forefront of AI infrastructure development, working with cutting-edge technologies and contributing to the advancement of deep learning frameworks.

As a key member of the R&D team, you'll collaborate with AI research scientists to accelerate the next generation of deep learning models. Your responsibilities will include developing custom architectures, performance tuning, and optimizing software components. The role offers exposure to both software systems and machine learning, providing an opportunity to understand PyTorch at a deeper level.

Meta offers a competitive compensation package, including a salary range of $56,250 to $173,000, plus bonus and equity opportunities. The company provides comprehensive benefits including medical, dental, and vision insurance. Working at Meta means being part of a team that's shaping the future of social technology and pushing the boundaries of what's possible in AR/VR.

The ideal candidate will have experience with GPU architectures, kernel frameworks like Triton or CUDA, and compiler optimizations. Knowledge of Large Language Models (LLM) serving frameworks is a plus. You'll be working in Menlo Park, CA, contributing to Meta's mission of connecting people and building immersive experiences that go beyond traditional digital connections.

This is an excellent opportunity for someone who wants to work at the intersection of systems and machine learning, making significant contributions to the AI infrastructure that powers Meta's various platforms and future technologies.