CUDA Engineer

This is a fully remote position, open to applicants in United States.

📋 Description

• Evaluate and enhance GPU kernels for improved performance, efficiency, and hardware utilization.

• Utilize profiling metrics like L2 cache hit rate, L2 throughput, occupancy, and similar indicators to inform kernel enhancements.

• Examine GPU kernel implementations and pinpoint bottlenecks without needing extensive knowledge of the underlying algorithms.

• Write, modify, and understand C++17, Python, and GPU programming code.

• Leverage expertise in CUDA, HIP, shader programming, or other kernel programming to boost performance results.

• Clearly document optimization decisions, indicating when specific profiling metrics are or are not beneficial.

⛳️ Requirements

• Must be available to work a minimum of 20 hours per week.

• Proficient in core C++ features up to C++17.

• Basic knowledge of Python and Git is required.

• Proficient in at least one GPU programming model, such as CUDA, HIP, Slang, HLSL, GLSL, or comparable kernel programming.

• A minimum of 1 year of professional or graduate-level research experience involving GPUs.

• Strong grasp of GPU profiler performance metrics and their application in kernel optimization.

• Capability to optimize GPU kernels without needing in-depth prior knowledge of every algorithm.

• Experience with CUDA, HIP, CUDA C++ Core Libraries, inline PTX assembly, or tensor core-level optimization is advantageous.

• Familiarity with optimizing kernels for NVIDIA Blackwell hardware is a plus.

• Knowledge of NSight Compute is an additional benefit.

• Prior experience with GPU hardware organizations such as NVIDIA, AMD, or Qualcomm is a plus.

• Contributions to open-source projects related to GPU kernel optimization are a plus.

🏝️ Benefits

• Competitive salary and compensation package.

• Opportunities for professional growth and development.

• Collaborative and innovative work environment.

• Access to the latest technology and tools in GPU programming.