GPU-Accelerated Sparse Linear Algebra Engineer

NVIDIA Corporation; Santa Clara (CA); On-site; USD 184,000 - 288,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA Corporation is seeking software engineers to design, implement, and optimize high-performance sparse linear algebra kernels for cuSPARSE and cuDSS. You’ll work on GPU-accelerated libraries, collaborating with product management and engineers to meet feature and performance goals across CUDA-enabled architectures.

Ideal candidates have 5+ years in HPC, strong C++ and parallel programming skills, and a solid background in sparse linear algebra primitives.

Qualifications

• PhD or MSc preferred in Computational Science & Engineering or related field.
• 5+ years in developing high-performance sparse linear algebra software in C++ and parallel environments.
• Strong fundamentals in floating point arithmetic and sparse primitives.

Responsibilities

• Design, implement and optimize scalable sparse linear algebra kernels for GPU architectures.
• Collaborate with library engineers, QA, and product teams to meet performance requirements.
• Contribute to roadmaps and improve library quality, performance, and maintainability.

Skills

C++

Parallel programming

Sparse linear algebra

Floating point arithmetic

Education

PhD or MSc in Computational Science/Engineering

Tools

CUDA

MPI

OpenMP

OpenACC

Pthreads

Job description

NVIDIA Corporation is seeking software engineers to design, implement, and optimize high-performance sparse linear algebra kernels for cuSPARSE and cuDSS. You’ll work on GPU-accelerated libraries, collaborating with product management and engineers to meet feature and performance goals across CUDA-enabled architectures.

Ideal candidates have 5+ years in HPC, strong C++ and parallel programming skills, and a solid background in sparse linear algebra primitives.