Workload Porting & Performance Engineer

About this role

OpenAI seeks a Workload Porting & Performance Engineer to evaluate new hardware platforms by porting AI workloads, analyzing performance across system components, and identifying bottlenecks. You'll optimize workloads for new hardware and provide insights to hardware and software teams to guide platform deployment decisions.

What you'll do

• Port benchmarks and real-world AI workloads onto new hardware platforms
• Evaluate performance across compute, memory, storage, and networking subsystems
• Identify and analyze performance bottlenecks and optimization opportunities
• Develop and execute performance experiments with profiling workflows
• Debug issues across software, runtime, and hardware layers
• Provide performance feedback to architecture, modeling, and engineering teams

What they're looking for

• Performance analysis and benchmarking
• System architecture knowledge (CPU, GPU, memory, I/O)
• Hardware platform porting and workload adaptation
• Profiling tools and performance debugging
• Root cause analysis across hardware/software boundaries
• Large-scale and distributed systems experience
• AI/ML workload optimization
• Low-level performance tools and instrumentation

Benefits

• Hybrid work model (3 days in office per week)
• Based in San Francisco, CA
• Relocation assistance offered