Build and maintain the software systems and control planes that run Google's Cloud TPU infrastructure. Own end-to-end features across the TPU software stack, which may include cluster and fleet management, power and resource orchestration, host and runtime software, hardware qualification, or fleet health and observability. Write production code, build automated validation pipelines, and improve telemetry and tooling to keep large-scale TPU clusters reliable and efficient for AI training and inference workloads.
Design, implement, and test software components, debug complex issues across distributed systems and hardware interfaces, and take features from initial design to production roll-out. Work closely with partner engineering teams across Google Cloud and data center infrastructure to understand customer and operational requirements.
Job description provided by Google.
View the original posting
