NVIDIA is seeking a CUDA Engineer to optimize deep learning frameworks for GPU acceleration. You will work on kernel optimization, memory management, and performance tuning for training and inference workloads on NVIDIA's latest GPU architectures.
Requirements
- 5+ years of CUDA/C++ programming experience
- Deep understanding of GPU architecture and parallel computing
- Strong background in computer architecture
- BS/MS/PhD in Computer Science or Electrical Engineering
Nice to Have
- Experience with TensorRT or Triton inference server
- Contributions to PyTorch or TensorFlow GPU backends
- Knowledge of mixed-precision training techniques
- Experience with NVIDIA's Hopper or Blackwell architectures
Stock grants in a leading semiconductor company
Employee stock purchase plan
On-site gym and cafeteria
Patent bonuses
Skills
CUDA C++ GPU Computing Deep Learning Performance Optimization Python
This listing was posted on 2026-02-26 and is almost certainly closed.
#J-18808-Ljbffr