Contract GPU Kernel Optimizer (CUDA/C++17)

Mercor; San Francisco (CA); On-site; USD 165,000 - 276,000; Full time

Job summary

Mercor seeks GPU kernel optimization experts for a freelance project in the San Francisco area. You will analyze, optimize, and reason about GPU kernels across modern hardware, squeezing performance with profiler-guided analysis.

Requirements include strong C++17, Python, and GPU programming experience (CUDA/HIP). Familiarity with NSight Compute and kernel profiling is a plus; work at least 20 hrs/week on contract basis.

Qualifications

• Fluent in core C++ features through C++17.
• Working knowledge of Python and Git.
• Fluent in at least one GPU programming model (CUDA, HIP, Slang, HLSL, GLSL).
• At least 1 year of GPU-related research or professional experience.
• Understanding of GPU profiler metrics and how to optimize kernels.

Responsibilities

• Analyze and optimize GPU kernels for performance, efficiency, and hardware utilization.
• Use profiler metrics (e.g., L2 cache hit rate, L2 throughput, occupancy) to guide kernel improvements.
• Review kernel implementations and identify bottlenecks with minimal context on algorithms.
• Write, modify, and reason about C++17, Python, and GPU programming code.
• Apply CUDA, HIP, shader programming, or related kernel programming expertise to improve outcomes.
• Document optimization decisions clearly, including usefulness of specific profiler metrics.

Skills

C++17

Python

Git

GPU profiling

CUDA

Tools

NSight Compute

Job description

Mercor seeks GPU kernel optimization experts for a freelance project in the San Francisco area. You will analyze, optimize, and reason about GPU kernels across modern hardware, squeezing performance with profiler-guided analysis.

Requirements include strong C++17, Python, and GPU programming experience (CUDA/HIP). Familiarity with NSight Compute and kernel profiling is a plus; work at least 20 hrs/week on contract basis.