Math Device Extension Libraries Intern - 2027
Core
Designing, developing, and optimizing GPU-accelerated mathematical libraries (cuBLASDx, cuSolverDx, etc.) and APIs for NVIDIA's accelerated computing platform.
Role type
Software engineering intern (high-performance computing)
Builds
GPU-accelerated math libraries and Dx APIs
Domain
High-performance computing (HPC) and AI
Deliverable
production ML models
Required skills
C++ programming, GPU programming (CUDA, OpenMP, OpenCL, SYCL), numerical algorithms (linear algebra, FFT, data compression), performance analysis, test building
Preferred skills
Compiler implementation, GPU/network hardware architecture, Python, LLMs, agentic workflows, numerical error analysis
Technologies
CUDA, cuBLASDx, cuSolverDx, cuFFTDx, cuRANDDx, nvCOMPDx, NPPDx, libmathdx, MLIR, Python, Numpy, JAX
Responsibilities
Prototype and develop numerical algorithms for high-performance math device extension libraries; Analyze performance of GPU implementations and find opportunities for improvements; Collaborate with team members to understand software use cases and requirements; Prototype tooling to transform math libraries' APIs into LLM-friendly representations
Seniority
Intern