Performance Engineer - Deep Learning
Core
Build and optimize libraries and tools (Transformer Engine, PyTorch, JAX) to accelerate Deep Learning model training and deployment on NVIDIA GPUs.
Role type
Senior IC performance engineer (Deep Learning systems)
Builds
Optimized Deep Learning frameworks and libraries for LLMs and AI applications
Domain
AI/ML infrastructure, GPU computing, High-Performance Computing
Deliverable
production ML models
Required skills
C++, Python, Parallel systems programming, GPU programming, Computer Architecture, Code Optimization, Operating Systems
Preferred skills
PyTorch, JAX, LLM architectures, CUDA, OpenAI Triton, CuTeDSL, Pallas, Multi-GPU/Multi-node profiling
Technologies
PyTorch, JAX, CUDA, OpenAI Triton, CuTeDSL, Pallas, cuBLAS, cuDNN, cuSOLVER, Transformer Engine, MLPerf
Responsibilities
Build and support Transformer Engine; Collaborate on systems research for model performance; Implement and optimize new Deep Learning models; Contribute to community benchmarks; Engage with open-source community and enterprise customers; Influence hardware and platform software design
Seniority
Mid-to-Senior, hands-on IC
