Senior HPC Middleware Engineer — Equity

NVIDIA; California (MO); On-site; USD 184,000 - 287,500; Full time

Job summary

NVIDIA is seeking a senior HPC software engineer to design and implement high-performance protocols and algorithms for NVIDIA networked supercomputers and datacenters. You will deliver production-grade software, work with leaders in ML, cloud computing, and system co-design, and help push the boundaries of performance and scalability.

The role emphasizes deep knowledge of networking, architecture, and OS concepts, with opportunities to contribute to MPI, RDMA, and open-source projects.

Qualifications

• 5 years of programming in C/C++
• 3 years in Linux environment and tools
• Deep knowledge of Networking Protocols InfiniBand, Ethernet
• Deep knowledge in computer architecture and operating systems
• Experience in performance optimizations
• PhD or equivalent experience in CS/EE/Math/Physics
• Knowledge in MPI and High-performance computing
• Knowledge in RDMA technology
• Open Source Software Contributor

Responsibilities

• Develop high-performance communication frameworks and applications running in production on large supercomputers and datacenters.
• Collaborate with chip design teams to co-design new hardware features and software APIs; push performance and scalability.
• Deliver production-grade software and contribute to performance optimization and software reliability.

Skills

C/C++ programming

Linux

Networking (InfiniBand, Ethernet)

Computer architecture

Performance optimization

MSc or equivalent experience

Education

MSc or equivalent

PhD or equivalent

Job description

NVIDIA is seeking a senior HPC software engineer to design and implement high-performance protocols and algorithms for NVIDIA networked supercomputers and datacenters. You will deliver production-grade software, work with leaders in ML, cloud computing, and system co-design, and help push the boundaries of performance and scalability.

The role emphasizes deep knowledge of networking, architecture, and OS concepts, with opportunities to contribute to MPI, RDMA, and open-source projects.