Senior Systems Software Engineer, Windows and Linux Enablement - DGX Station

NVIDIA•Santa Clara, CA

1d•$224,000 - $356,500

About The Position

DGX Station is NVIDIA’s next-generation personal AI supercomputer—a deskside workstation built on the NVIDIA Grace Blackwell GB300 Superchip with massive coherent CPU+GPU memory, designed to bring data-center-class AI capabilities directly to the desks of researchers, developers, and AI engineers. As NVIDIA brings DGX Station to a broad set of customers, we need an engineer who can own full-stack OS enablement—from firmware and drivers through OS integration to ensuring AI applications run seamlessly on day one, with a primary focus on Windows and strong coverage of Linux. This is a hands-on, technically deep role where you will be the go-to engineer for making DGX Station a first-class Windows platform while also driving its Linux bring-up and validation. You will work across NVIDIA’s GPU driver, CUDA, firmware, BMC, and AI software teams, collaborate closely with Microsoft and ODM/OEM partners, and ensure that developers and enterprise customers have a polished, production-ready experience on DGX Station across both operating systems.

Requirements

BS or MS in Computer Science, Electrical Engineering, or related field (or equivalent experience) and 12+ yrs of confirmed experience in systems software engineering with deep expertise in Windows platform enablement, driver development, or OS integration, and proven hands-on experience bringing up Linux on new hardware platforms.
Strong hands-on experience with Windows internals: kernel-mode drivers, ACPI, power management, Secure Boot, UEFI, WDM/WDF driver frameworks, and the WHQL certification process.
Solid understanding of Linux platform enablement: kernel modules, device tree / ACPI on Arm, systemd, initramfs, dkms, and packaging for Ubuntu / DGX OS.
Experience with GPU driver stack, display drivers, or compute drivers on Windows and/or Linux.
Experience enabling hardware platforms—bring-up, driver integration, validation, and certification for shipping products on Windows and Linux.
Strong debugging and root-cause analysis skills across firmware, driver, and OS boundaries.
Comfortable with WinDbg, kernel debugging (kd, kgdb/crash), crash dump analysis, ftrace/ETW, and performance profiling tools.
Ability to work across organizational boundaries—coordinating with GPU driver, CUDA, firmware, BMC, and AI software teams as well as external partners (Microsoft, ODM/OEMs).
Proficiency in C/C++ and Python.

Nice To Haves

Familiarity with DirectX, WDDM, DRM/KMS, and GPU compute APIs is a strong plus.
Experience with Arm architecture is a plus.
Experience with Windows on Arm platforms—driver enablement, performance optimization, or application compatibility on Arm-based Windows devices.
Hands-on experience with CUDA, TensorRT, or AI/ML frameworks on Windows and Linux—especially on Arm + NVIDIA GPU systems.
Prior experience working with OEM/ODM partners or silicon vendors on Windows and Linux platform certification for workstation- or server-class hardware.
Track record shipping workstation or server hardware products—from bring-up through general availability—with both Windows and Linux support.
Experience with BMC, Redfish, out-of-band management, or platform manageability software on high-end workstations or servers.
Experience with GPU-accelerated applications: AI training and inference, content creation tools, or scientific computing on Windows and Linux.

Responsibilities

Own end-to-end Windows enablement for DGX Station—driving the platform from initial bring-up on Windows through WHQL certification to customer-ready shipping quality.
Drive Linux bring-up and continuous enablement for DGX Station on DGX OS / Ubuntu, including kernel module integration, device tree and ACPI configuration, systemd services, initramfs, and dkms packaging.
Enable and validate BIOS/UEFI, BMC, and system-level firmware for Windows and Linux on the Grace (Arm) + Blackwell GB300 architecture.
Coordinate GPU driver, display driver, and compute driver bring-up and validation on Windows (WDDM, MCDM) and Linux (open-gpu-kernel-modules, DRM/KMS).
Ensure the CUDA toolkit, cuDNN, TensorRT, NCCL, and NVIDIA’s AI SDK stack are fully functional on DGX Station on both Windows and Linux.
Validate that NVIDIA AI applications—NIM microservices, NemoClaw, AI Workbench, and developer tools—run correctly on DGX Station across Windows and Linux.
Drive the overall test strategy for DGX Station on Windows and Linux: functional testing, stress testing, power/thermal validation, sleep/resume and S-state cycles, Windows Update and Linux kernel-upgrade compatibility, and long-duration reliability.
Be the primary technical interface with Microsoft (Windows on Arm, WHQL, driver signing) and ODM/OEM partners shipping DGX Station.
Profile and optimize system performance—boot time, GPU compute throughput, NVLink-C2C and memory bandwidth utilization, power efficiency, and thermal behavior.
Create and maintain platform documentation for DGX Station on Windows and Linux: bring-up guides, known issues, driver compatibility matrices, recovery and re-imaging procedures, and developer setup instructions.