Senior GPU Performance Architect & Optimization Lead

Jobtailor; California (MO); On-site; USD 150,000 - 210,000; Full time

Job summary

Jobtailor is seeking a seasoned Systems Performance Engineer to drive GPU/HPC/ML workload performance for CSP/hyperscale deployments. You will profile workloads, analyze bottlenecks, and champion optimizations across NVIDIA CUDA, NCCL, drivers, and firmware teams.

Responsibilities include coordinating with engineering teams, defining test strategies, and ensuring measurement tools remain up-to-date with the latest architectures and platforms.

Qualifications

• 15+ years of experience in systems performance engineering.
• BS or MS in Computer Science, Computer Engineering, or related field (or equivalent).
• Proficiency with GPU workload profiling tools and instrumentation.
• Understanding of distributed training performance dynamics.
• Statistical methods for performance analysis and AB comparisons at scale.
• Knowledge of how software stack affects performance: drivers, memory, scheduling, firmware powers.
• Strong data analysis and visualization skills (Python, pandas).
• Ability to communicate findings to technical and executive audiences.
• Proven track record influencing multiple engineering teams to prioritize performance improvements.

Responsibilities

• Drive performance characterization with engineering teams for GPU/HPC/ML workloads.
• Synthesize CSP performance feedback and champion optimization priorities across CUDA, NCCL, drivers, and firmware teams.
• Ensure performance tools (e.g., nsight tools, DCGM) are updated and validated for latest NVIDIA rack-scale systems and architectures.
• Collaborate with CSPs to align profiling and benchmarking tooling with platform capabilities.
• Define test strategies and tooling requirements for performance validation for internal and customer approval.

Skills

GPU profiling

Python

Pandas

Data visualization

Statistical analysis

Cross-team collaboration

Performance optimization

Executive communication

Education

BS in Computer Science or Engineering

MS in Computer Science or Engineering

Tools

Nsight Systems

Nsight Compute

DCGM Metrics

Job description

Jobtailor is seeking a seasoned Systems Performance Engineer to drive GPU/HPC/ML workload performance for CSP/hyperscale deployments. You will profile workloads, analyze bottlenecks, and champion optimizations across NVIDIA CUDA, NCCL, drivers, and firmware teams.

Responsibilities include coordinating with engineering teams, defining test strategies, and ensuring measurement tools remain up-to-date with the latest architectures and platforms.