Used Tools & Technologies
Not specified
Required Skills & Competences ?
Docker @ 3 Linux @ 2 Communication @ 3 Networking @ 3 Debugging @ 2 API @ 3Details
NVIDIA’s Data-Center-on-a-Chip (DOCA) Hyperscale Engineering team is seeking a part-time Software Research Engineer—ideally a PhD candidate in Computer Science or Electrical Engineering—to help prototype, evaluate, and extend our software-defined networking stack on the BlueField DPU / CX SuperNICs. You’ll join a small, collaborative group of SDK developers, firmware engineers, and DPU architects to gain hands-on experience with groundbreaking networking technologies and contribute to our next-generation data-center platform.
Are you interested in direct mentorship from NVIDIA architects and engineers at the forefront of DPU SoC technology? Join us to gain hands-on experience optimizing and scaling network applications in hardware-accelerated environments. You’ll have the opportunity to shape both an open SDK (DOCA) and internal research initiatives, all while enjoying competitive hourly compensation and access to NVIDIA’s world-class technical resources.
Responsibilities
- Collaborate with DOCA/DPDK engineers to prototype new features and benchmarks for our BlueField DPU platform
- Develop, test, and document C/C++ applications that leverage hardware-accelerated networking primitives
- Build performance measurement tools and run experiments to analyze throughput, latency, and offload efficiency
- Participate in design reviews, code walkthroughs, and team discussions on architecture and APIs
- Write clear technical reports and present findings to the DOCA team and academic advisors
- Aid in customer or partner demos, preparing scripts and helping with live virtualization or RDMA showcases
Requirements
- Pursuing an MS degree in Computer Science, Computer Engineering, or Electrical Engineering; PhD preferred
- Strong C and/or C++ programming skills, with coursework or projects in systems, networking, or embedded software
- Familiarity with Linux development tools, build systems (CMake, Make), and debugging tools (gdb, perf)
- Solid understanding of basic networking concepts (TCP/IP, UDP, packet processing)
- Excellent written and verbal communication skills
- Self-motivated and well organized, able to balance academic and part-time work responsibilities
- Position may be remote or hybrid (onsite at NVIDIA’s US offices) depending on location and availability
Ways to Stand Out From the Crowd
- Prior experience with DPDK or kernel-bypass networking libraries
- Exposure to virtualization (KVM, Docker) or container networking
- Hands-on projects involving RDMA, SmartNICs, or hardware offloads
- Background in performance analysis or low-level optimization
Benefits
- Competitive hourly compensation ($18 - $71 USD per hour)
- Eligibility for Intern benefits
- Mentorship from leading industry engineers
- Access to NVIDIA’s cutting-edge technical resources
NVIDIA values diversity and is an equal opportunity employer, committed to fostering a diverse work environment free from discrimination.