Senior Math Libraries Engineer, Sparsity in AI

What you'll do

• Design and develop a C++-based system to accelerate unstructured sparsity for DL and HPC on NVIDIA GPUs.
• Enable the system for high-level languages and frameworks including Python and PyTorch.
• Evaluate and improve system performance using real-life applications.
• Write effective, well-tested code to improve library quality, performance, and maintainability for production use.
• Develop solutions for sparse tensor computations and domain-specific language specifications for sparse storage formats.
• Implement on-demand code generation for accelerated computing.
• Contribute to technical roadmaps by defining feature and performance requirements with internal and external partners.

What we're looking for

• BS, MS, or PhD degree in Computer Science, Applied Math, or a related field (or equivalent experience).
• 6+ years of experience developing, debugging, and optimizing high-performance software using C++ and parallel programming.
• Experience with sparse linear algebra applications using CUDA, MPI, OpenMP, or equivalent technologies (preferred).
• Experience with domain-specific language design and compiler optimizations, specifically sparse compilers like MLIR or TACO.
• Excellent programming skills in C++, Python, and CUDA.
• Strong collaboration, communication, and documentation habits, ideally within a globally distributed organization (preferred).
• Strong understanding of sparse computations in AI and HPC (preferred).
• Good understanding of LLMs, Deep Learning methods, and frameworks (preferred).