Senior Math Libraries Engineer - Sparsity in AI
Core
Design and develop C++-based systems to simplify and accelerate unstructured sparsity computing for Deep Learning and HPC on NVIDIA GPUs.
Role type
Senior IC software engineer (high-performance computing & sparse linear algebra)
Builds
GPU-accelerated libraries, SDKs, and domain-specific language (DSL) specifications for sparse storage formats
Domain
High-performance computing, Deep Learning, GPU acceleration, Sparse Linear Algebra
Deliverable
production ML models
Required skills
C++, CUDA, Python, parallel programming, sparse tensor computations, domain-specific language design, compiler optimizations
Preferred skills
LLMs, Deep Learning frameworks, low-level GPU performance optimization, numerical linear algebra methods, CI/CD systems
Technologies
C++, Python, CUDA, PyTorch, MLIR, TACO, MPI, OpenMP, JIRA
Responsibilities
Design and develop C++ systems for unstructured sparsity in DL and HPC; Enable the system in Python and PyTorch; Evaluate and improve system performance on real-life applications; Write effective and well-tested code for production use; Collaborate with product management and partners on feature and performance requirements
Seniority
Senior, hands-on IC
