Meta is seeking an experienced software engineer to join their Accelerator Solutions & Technologies group, focusing on developing cutting-edge collective communications software libraries and optimizing distributed AI/ML workloads. This role presents a unique opportunity to work with Meta's next-generation, super-cluster AI/ML platforms at an unprecedented scale.
The position involves working with a highly skilled engineering team and collaborating with cross-functional and international partners. You'll be at the forefront of developing and optimizing Meta's accelerator programs, contributing to simulation platforms, and ensuring system designs meet rigorous performance standards.
Key responsibilities include implementing simulation models for Meta's Accelerator ASICs, developing the collective communications library, and supporting hardware acceleration techniques. You'll work closely with hardware and firmware teams, using instruction set simulators to define performant firmware and analyze results for optimization.
The ideal candidate should have a strong educational background in Computer Science or related fields, with at least 2 years of experience in both C++ and Python development. Knowledge of AI/HPC systems, familiarity with frameworks like PyTorch and CUDA, and understanding of hardware requirements are crucial for success in this role.
This is an exceptional opportunity to impact the performance and efficiency of some of the world's largest scale clusters. You'll be working at Meta, a company at the forefront of social technology and immersive experiences, helping to shape the future of digital connection beyond traditional constraints.
The role offers competitive compensation ranging from $56,250 to $173,000 annually, plus bonus, equity, and comprehensive benefits. Based in Menlo Park, CA, you'll be part of Meta's innovative culture that values diversity, equality, and inclusion.