Senior Math Libraries Engineer - Direct Sparse Solvers

at Nvidia

USD 184,000-356,500 per year

SENIOR

✅ On-site

Tech Stack
Tag name is followed by "@" symbol and proficiency level value. About proficiency levels:

1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;

3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;

7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;

10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.

AI @ 4 Agile @ 4 Algorithms CI/CD @ 4 CUDA @ 4 Communication @ 7 Debugging @ 6 GPU @ 4 Jira @ 4 LLM MPI @ 4 Mathematics @ 4 Parallel Programming @ 6 Performance Optimization @ 4 Product Management @ 4 Project Management @ 4

Details

NVIDIA is seeking software engineers to join the development efforts for cuDSS, a CUDA library for direct solvers for sparse linear systems. The team develops GPU-accelerated libraries and SDKs used in scientific and engineering simulations, data analytics, AI, computer-aided engineering, electronic design automation, optimization, quantum chemistry, autonomous vehicles, large language models, and other applications.

The role involves crafting algorithms and kernels for direct sparse solvers and advancing accelerated computing technologies.

Responsibilities

Design, implement, and optimize direct sparse solvers for existing and future GPU architectures.
Work with library engineers, QA engineers, and interns across library development, including design, implementation, testing, release, and support.
Collaborate with product management and internal and external partners to understand feature and performance requirements and contribute to technical roadmaps.
Identify and implement opportunities to improve the quality, performance, and maintainability of sparse linear algebra libraries through re-architecting and innovative software development practices.

Requirements

PhD or MSc degree in Computer Science, Computational Science and Engineering, Applied Mathematics, or a related science or engineering field, or equivalent experience.
At least 5 years of experience developing, debugging, and optimizing high-performance numerical software using C++ and parallel programming.
Experience with CUDA, MPI, OpenMP, OpenACC, pthreads, or equivalent technologies is preferred.
Strong foundations in floating-point arithmetic, numerical analysis, and sparse linear algebra primitives such as matrix-vector products, matrix-matrix products, and triangular solves.
Experience developing, maintaining, and testing scientific computing libraries.
Strong collaboration, communication, and documentation skills.

Preferred Qualifications

Familiarity with direct solver techniques, including reordering, multifrontal factorizations, supernodal factorizations, numerical pivoting strategies, and iterative refinement.
Knowledge of CPU and/or GPU hardware architecture and low-level GPU performance optimization.
Experience adopting and advancing modern software engineering methods, including CI/CD systems, project management tools such as JIRA, and AI agents.
Understanding of large-scale computing technologies such as PDE solvers, eigenvalue solvers, and time-domain simulation methods, including CFD and FEA.
Experience working in a globally distributed and agile organization.

Benefits

Equity and employee benefits.
NVIDIA is an equal opportunity employer committed to an inclusive work environment.
Applications will be accepted at least until June 13, 2026.
NVIDIA uses AI tools in its recruiting processes.