Software Engineer, Tensor Processing Units Compiler

Google is seeking a Software Engineer for their Tensor Processing Units (TPUs) Compiler team. This role is critical in developing the compiler that enables TPUs, Google's custom-designed processor, to accelerate machine learning and scientific computing workloads for both internal and external Cloud customers. The position offers opportunities to work on various levels of the compiler stack, including LLVM and MLIR middle-end.

Key responsibilities include contributing to the TPU compiler, supporting new workloads, optimizing for new models and characteristics, and supporting new TPU hardware across multiple generations. The ideal candidate will have experience with CUDA C++ application development, compiler technologies, and machine learning algorithms.

The role requires collaboration with TPU users to improve performance and efficiency, as well as with hardware designers for co-designing future processors. The engineer will also investigate high-level representations for programming large-scale, distributed, and heterogeneous systems.

Google Cloud, the division this role supports, is at the forefront of enabling digital transformation for organizations worldwide. It offers enterprise-grade solutions leveraging Google's cutting-edge technology on the cleanest cloud in the industry.

This position offers an exciting opportunity to work on advanced technology that impacts billions of users and pushes the boundaries of machine learning acceleration. The ideal candidate will be versatile, display leadership qualities, and be enthusiastic about tackling new problems across the full stack as Google continues to innovate in the field of AI and machine learning.