Senior GPU Performance Engineer for AI/ML

Google; United Kingdom; On-site; GBP 120,000 - 180,000; Full time

Job summary

Google is seeking software engineers to advance next-generation AI and ML systems. You will work on tasks spanning information retrieval, distributed computing, large-scale system design, security, and AI.

Role emphasizes leadership, versatility, and the ability to switch teams as projects evolve. The ML-Omega team focuses on optimizing GPU systems for scalable ML workloads across Google's infrastructure.

Qualifications

• Bachelor's degree or equivalent practical experience is required.
• 5 years of software development experience in one or more programming languages.
• 5 years of experience with data structures and algorithms.
• 5 years of ML algorithms/tools, AI, DL, or NLP.
• 3 years of testing/maintaining/launching software; 1 year in design/architecture.

Responsibilities

• Identify LLM training and serving benchmarks representative of Google production and ML community.
• Collaborate with product teams to onboard new LLM models and enable large-scale training/serving.
• Run architecture-level simulations on GPU designs and perform roofline analysis.
• Analyze performance metrics to identify bottlenecks and implement fleet-wide improvements.
• Run GPU benchmarks using internal and external tools.

Skills

Software development

Data structures & algorithms

Machine learning

Testing/maintenance/launching

Software design & architecture

Education

Bachelor's degree or equivalent practical experience

Master's or PhD in CS or related field

Tools

CUDA

Triton kernels

Palace/Mosaic kernels

XLA

Job description

Google is seeking software engineers to advance next-generation AI and ML systems. You will work on tasks spanning information retrieval, distributed computing, large-scale system design, security, and AI.

Role emphasizes leadership, versatility, and the ability to switch teams as projects evolve. The ML-Omega team focuses on optimizing GPU systems for scalable ML workloads across Google's infrastructure.