Software Engineer III, Performance
Mountain View, CA, USA · Full-time · Posted Oct 22, 2026
- Experience
- 2 years of software development and 2 years with computer architecture, performance analysis, and performance modeling.
- Education
- Bachelor’s degree or equivalent practical experience.
About the job
Optimize open-source models for Google TPUs, develop scalable performance tooling and recipes, and identify performance issues in current and unseen models.
What you'll do
- Write product or system development code and manually optimize models using techniques such as sharding, quantization, and custom kernels.
- Turn optimization principles into scalable solutions, including cost models and agentic tooling.
- Improve out-of-the-box software performance, open-source customer tooling, and maintain model benchmarks and trend analysis.
What they ask for
- Bachelor’s degree or equivalent practical experience.
- 2 years of software development experience in one or more programming languages.
- 2 years of experience with computer architecture, performance analysis, and performance modeling.
- Experience experimenting with AI agents or LLM-powered agentic workflows, and experience in systems engineering.
Nice to have
- Master's degree or PhD in Computer Science or related technical fields.
- Experience with low-level accelerator programming, machine learning, compiler optimization, code generation, runtime systems, or ML accelerator performance tuning.
Pay
$147000 - $210000 (USD) base salary
Tools
TPUs, CUDA, Triton, Pallas, OpenXLA, MLIR
AI focus
Develop principles from manual model optimizations that can be automated via agentic AI, including cost models and agentic tooling; experience with AI agents or LLM-powered agentic workflows is required.
Model-written summary · the original posting is authoritative