Location
toronto, on, Canada
Posted
June 14, 2026
Job Description
Join NVIDIA as a GPU Performance Engineer focused on neural reconstruction and elevate the future of computing. Your expertise in CUDA and PyTorch will be key in optimizing sophisticated 3D systems.
This role involves enhancing GPU performance for innovative workloads like Gaussian Splatting and neural reconstruction. As a highly skilled engineer, you'll work closely with hardware while improving training and rendering workflows. Collaboration with researchers and engineers will ensure that prototypes transform into maintainable production-quality code.
Key Responsibilities:
• Profile neural reconstruction workflows to identify bottlenecks
• Optimize CUDA and PyTorch performance for complex workloads
• Analyze GPU performance using various profiling tools
• Develop benchmarks and profiling workflows for performance validation
• Translate Python bottlenecks into efficient CUDA/C++ implementations
Requirements:
• 12+ years of experience in a relevant field
•...
This role involves enhancing GPU performance for innovative workloads like Gaussian Splatting and neural reconstruction. As a highly skilled engineer, you'll work closely with hardware while improving training and rendering workflows. Collaboration with researchers and engineers will ensure that prototypes transform into maintainable production-quality code.
Key Responsibilities:
• Profile neural reconstruction workflows to identify bottlenecks
• Optimize CUDA and PyTorch performance for complex workloads
• Analyze GPU performance using various profiling tools
• Develop benchmarks and profiling workflows for performance validation
• Translate Python bottlenecks into efficient CUDA/C++ implementations
Requirements:
• 12+ years of experience in a relevant field
•...