Machine Learning Hardware Architect, Google Cloud

Google Cloud is seeking a Machine Learning Hardware Architect to shape the future of AI/ML hardware acceleration, specifically focusing on TPU (Tensor Processing Unit) technology. This role combines advanced computer architecture with machine learning expertise, requiring deep understanding of both hardware design and ML frameworks.

The position involves developing cutting-edge TPU architecture that powers Google's most demanding AI/ML applications. You'll work on optimizing accelerator performance under power constraints, developing transaction-level models, and streamlining host-accelerator interactions. The role requires collaboration with various teams including XLA compiler, Platforms performance, and system design teams.

As part of the ML, Systems, & Cloud AI (MSCA) organization, you'll contribute to the infrastructure supporting all Google services and Google Cloud. The work impacts billions of users through products like Google Cloud's Vertex AI and various Google services. The role offers competitive compensation including base salary, bonus, equity, and comprehensive benefits.

The ideal candidate should have strong experience in computer architecture and hardware design, coupled with software development skills in C++ or Python. Knowledge of ML frameworks like TensorFlow and PyTorch is highly valued, as is understanding of the ML market and technological trends. This position offers an opportunity to work at the intersection of hardware architecture and machine learning, driving innovation in AI acceleration technology.

Working at Google provides exposure to cutting-edge technology and the chance to impact global-scale systems. The role combines technical depth with broad system-level thinking, requiring collaboration across multiple specialized teams to deliver comprehensive solutions for AI hardware acceleration.