CUDA / GPU Performance Engineer – Kernel Optimization

This is a fully remote position, open to applicants in India, +5 more states.

📋 Description

• Analyze and enhance CUDA kernels to improve throughput, latency, and hardware utilization.

• Profile GPU workloads to pinpoint compute, memory, synchronization, and execution bottlenecks.

• Design and implement specific kernel optimization techniques.

• Refactor C++ and CUDA codebases to increase performance, maintainability, and portability.

• Assess kernel performance across various GPU architectures and hardware generations.

• Create or modify shader and compute workflows utilizing GLSL and WebGPU.

• Utilize GPU profiling tools to verify enhancements and evaluate performance.

• Document optimization strategies, benchmarks, results, and performance improvements.

• Provide technical insights during GPU architecture and performance-design discussions.

• Investigate new GPU programming methodologies and implement applicable improvements.

⛳️ Requirements

• Extensive professional experience in CUDA programming and GPU kernel optimization.

• Advanced skills in C++, preferably in high-performance or systems programming contexts.

• Demonstrated experience in profiling and optimizing GPU workloads for enhanced performance.

• Practical experience with GPU profiling tools like NVIDIA Nsight or similar alternatives.

• Strong comprehension of GPU architecture, memory hierarchy, parallel execution, and synchronization.

• Experience in evaluating performance across diverse GPU hardware generations.

• Practical experience with GLSL and/or WebGPU for shader or compute development.

• Ability to clearly document performance observations and technical decisions in English.

🏝️ Benefits

• Competitive salary and performance-based bonuses.

• Health, dental, and vision insurance plans.

• Opportunities for professional development and training.

• Flexible work hours and remote work options.

• Engaging and collaborative work environment.