GPU Performance Engineer

Job Summary

Focused on optimizing neural reconstruction workflows, the full-time GPU Performance Engineer will enhance GPU performance using CUDA and PyTorch, collaborating with cross-functional teams in a remote environment.

Key responsibilities

• Profile and identify bottlenecks in end-to-end neural reconstruction workflows across various stages
• Improve performance for Gaussian Splatting and neural reconstruction workloads, ensuring quality and reliability
• Analyze GPU performance using profiling tools and develop benchmarks to validate optimizations

Required qualifications

• BS, MS, PhD, or equivalent experience in a relevant field with 12+ years of experience
• Strong programming skills in Python and C++
• Hands-on experience with PyTorch or similar tensor/autograd frameworks
• Experience optimizing GPU-accelerated workloads using CUDA and related programming approaches
• Practical experience with performance analysis and profiling of CPU/GPU bottlenecks