Sparsity in AI: GPU-Accelerated Math Libraries Engineer

NVIDIA Gruppe; Santa Clara (CA); On-site; USD 184,000 - 287,500; Full time

Benefits offered by this job

Equity

Benefits

Diversity and inclusion programs

Job summary

NVIDIA Gruppe is seeking experienced software engineers in Santa Clara, California, to design and develop systems that simplify computing for unstructured sparsity in deep learning and high-performance computing. The role involves working with C++ and collaborating with product management to understand requirements while contributing to the design of GPU-accelerated libraries.

Ideal candidates have 6+ years of experience with high-performance software, strong skills in C++, Python, and CUDA, and are familiar with sparse computations. The position offers a competitive salary based on experience and eligibility for equity and benefits.

Qualifications

• 6+ years of experience in developing, debugging and optimizing high-performance software using C++.
• Experience with domain-specific language design and compiler optimizations, specifically sparse compilers.
• Strong collaboration and documentation habits, ideally in a distributed organization.

Responsibilities

• Design and develop a C++-based system for unstructured sparsity in deep learning.
• Enable the system in languages like Python and PyTorch.
• Evaluate and improve performance of the system on real applications.

Skills

C++ programming

Python programming

CUDA

Parallel programming

Collaboration and communication

Education

BS, MS or PhD in Computer Science, Applied Math, or related field

Tools

JIRA

MPI

OpenMP

Job description

NVIDIA Gruppe is seeking experienced software engineers in Santa Clara, California, to design and develop systems that simplify computing for unstructured sparsity in deep learning and high-performance computing. The role involves working with C++ and collaborating with product management to understand requirements while contributing to the design of GPU-accelerated libraries.

Ideal candidates have 6+ years of experience with high-performance software, strong skills in C++, Python, and CUDA, and are familiar with sparse computations. The position offers a competitive salary based on experience and eligibility for equity and benefits.