Senior System Software Engineer - vGPU Team at Nvidia, Pune
Open. Posted .
GPU virtualization system software across kernel drivers, hypervisors and virtual machines, with architecture ownership, performance optimization and technical mentorship. The role is in Pune.
Applying opens the employer's page. Want jobs like this sent to you each morning? Open Elip.
| Location | Pune, Maharashtra |
|---|---|
| Work mode | On-site |
| Experience | 12–13 years |
| Pay | Not stated by the employer |
| Role | Embedded Engineer |
| Skills asked for | C, C++, Operating System Internals, Kernel Programming, Device Drivers, Memory Management, Concurrency, GPU Architecture |
| Posted | 16 Nov 2026 |
Elip's summary
Builds system software for a semiconductor and GPU computing company. Develops GPU virtualization across kernel-mode drivers, user-mode components, virtual machines, and hypervisors, applying C, C++, and operating system internals. Integrates VMware ESXi and KVM, and optimizes GPU resource sharing and performance.
The job, in the employer's words
NVIDIA is seeking an experienced Principal System Software Engineer to lead the design, development, and optimization of next-generation GPU virtualization technologies. The role focuses on enabling efficient, secure, and scalable sharing of NVIDIA GPUs across virtual machines and virtualized computing environments.
As a technical leader, you will work across GPU hardware, kernel-mode drivers, user-mode software, hypervisors, and virtualization platforms to deliver high-performance graphics and GPU computing capabilities. You will drive architecture and implementation, resolve complex system-level issues, and collaborate with cross-functional engineering teams to bring new GPU features to market.
What you will be doing:
Architecture and technical leadership: Define system software architecture for GPU virtualization, including GPU sharing, resource management, isolation, and virtualization support across multiple platforms.
Kernel and user-mode development: Design, implement, debug, and optimize system software across kernel-mode GPU drivers and user-mode components. Drive new GPU virtualization features from architectural design through implementation, testing, validation, and production release.
GPU virtualization: Develop and improve technologies that enable virtual machines to access GPU resources efficiently, including virtual GPU device management and GPU resource scheduling.
Performance optimization: Analyze GPU utilization, latency, throughput, memory usage, and virtualization overhead to improve performance for graphics and compute workloads.
Hypervisor integration: Work with virtualization technologies such as VMware ESXi, KVM, and other supported hypervisor environments to integrate GPU virtualization features.
System-level debugging: Investigate complex issues involving GPU drivers, operating systems, hypervisors, GPU hardware, memory management, concurrency, and system stability.
Cross-functional collaboration: Partner with GPU architecture, hardware, firmware, driver, QA, performance, and platform teams to resolve technical challenges.
Technical mentorship: Mentor engineers, conduct design and code reviews, establish engineering best practices, and influence technical decisions across teams.
What we need to see:
Bachelor’s or Master’s degree in Computer Science, Computer Engineering, Electrical Engineering, or a related field, or equivalent experience.
12+ years of experience in system software development, preferably at a senior staff, principal, or equivalent technical leadership level.
Strong proficiency in C and C++, with experience developing complex, low-level system software. Experience in Rust programming will be an added advantage
Deep understanding of operating system internals, kernel programming, device drivers, memory management, synchronization, interrupts, and concurrency.
Experience with GPU architecture, GPU drivers, graphics or compute pipelines, and hardware–software interaction.
Strong knowledge of virtualization concepts, including virtual machines, hypervisors, device virtualization, and I/O virtualization.
Experience with Linux kernel internals and debugging tools; Windows driver development experience is an advantage.
Demonstrated ability to lead complex technical projects, influence architecture decisions, and collaborate across engineering teams.
Proficiency in the use of AI tools for end to end SDLC activities
Ways to stand out from the crowd:
Hands-on experience with NVIDIA vGPU, GPU virtualization, or GPU passthrough technologies. VMware vSphere/ESXi, KVM, QEMU, VFIO, SR-IOV, or mediated device frameworks.
Knowledge of GPU memory virtualization, DMA, IOMMU, PCIe, MMIO, and GPU scheduling and experience with GPU isolation, multi-tenant security, resource allocation, virtualization performance optimization, strong debugging skills using kernel debuggers, crash dumps, tracing, profiling, and hardware diagnostics.
Familiarity with GPU compute technologies such as CUDA and graphics APIs such as DirectX, Vulkan, or OpenGL.
A track record of delivering complex system software across multiple hardware generations and operating system platforms.
About Nvidia
Headquartered in Santa Clara, founded in 1993.
- People in Elip's index who work here
- 56,179
- Typical time at Nvidia
- 3.9 years
- Typical career length so far
- 13.3 years
Largest teams
- Engineering24,591 people
- Information Technology3,955 people
- Operations2,417 people
- Research2,026 people
Where people here worked before
- Intel3,100 people
- Mellanox Technologies1,320 people
- Qualcomm1,049 people
- Google881 people
- AMD866 people
Where people went next, since 2025
- Google180 people
- Meta172 people
- Apple114 people
- AMD78 people
- Amazon76 people
Totals from public professional profiles in Elip's index; any group under ten people is left out.
Similar open jobs
About this page
Elip lists jobs that employers publish on their own careers sites, never copies from job boards. This one comes from Nvidia's listing on Workday, posted 16 Nov 2026. The summary is Elip's; the description is the employer's. When the job closes, this page says so.