✨ About The Role
- The role involves co-designing future hardware for programmability and performance with hardware vendors.
- Engineers will assist hardware vendors in developing optimal kernels and integrating support into the compiler.
- The position requires developing performance estimates for critical kernels across different hardware configurations.
- Communication and coordination with internal and external partners will be a key responsibility.
- The role will influence the roadmap of hardware partners to optimize them for OpenAI's workloads.
âš¡ Requirements
- The ideal candidate will have 4+ years of industry experience, particularly in harnessing compute at scale or building semiconductors.
- A strong background in software/hardware co-design is essential for success in this role.
- Candidates should possess a deep understanding of GPU and/or other AI accelerators.
- Experience with CUDA or a related accelerator programming language is highly desirable.
- The successful individual will be able to collaborate effectively with machine learning engineers, kernel writers, and compiler developers.