Machine Learning Compiler Software Engineer, TPU Horizontal Scaling

Google is seeking a Machine Learning Compiler Software Engineer for their TPU Horizontal Scaling team. This role involves developing the Accelerated Linear Algebra (XLA) TPU/GPU parallelizing compiler used to partition, optimize, and run large-scale machine learning models across multiple TPU/GPU accelerators for internal and external customers. The team's software stack includes the XLA Single Program Multiple Data (SPMD) partitioner, collective and scheduling optimizations, and code generation.

Key responsibilities include:

1. Writing product or system development code
2. Participating in or leading design reviews with peers and stakeholders
3. Contributing to a compiler that scales-out machine learning models across accelerators
4. Conducting static and runtime performance analysis of large-scale production models
5. Designing and implementing performance optimizations and critical features

The ideal candidate should have a Bachelor's degree or equivalent practical experience, with typically 2 years of experience in software development and data structures or algorithms. Preferred qualifications include a Master's or PhD in Computer Science or related field, experience in Machine Learning and High Performance Computing (HPC), distributed scale optimization, C++ programming, compilers, and debugging concurrent/parallel computations.

Google Cloud accelerates digital transformation for organizations, leveraging cutting-edge technology and tools for developers. This role offers the opportunity to work on critical projects, with the potential to switch teams and projects as the fast-paced business evolves. Google is committed to diversity, equality, and creating a culture of belonging for all employees.