NVIDIA is hiring a senior system software engineer to design and implement features in the CUDA driver for Microsoft Windows. The role requires strong C/C++ skills, experience with operating system interfaces, multithreaded programming, system architecture, and working on large codebases; Windows driver and kernel-mode experience, GPU/graphics knowledge, and WDDM experience are listed as differentiators. The posting includes salary ranges and is open through November 3, 2026.

Role focuses on analyzing end-to-end performance of large-scale AI workloads, designing and executing performance studies and benchmarks, using profiling and observability to diagnose bottlenecks, and driving optimizations with platform and GPU architecture teams. Requires BS (or equivalent), 12+ years' experience, strong C++ and Python skills, and foundations in operating systems, computer architecture, distributed systems, and performance engineering. Experience with CUDA, GPU systems, and deep learning frameworks (PyTorch, JAX/XLA) is highlighted.

NVIDIA seeks an engineer to design and implement VLSI tools and package design methodologies, developing flows, tools, and automation for multi-die packages and substrates. Requires a BS in CS/EE (or equivalent), 5+ years of package flow/methodology development experience, strong Python/C++ skills, AI/ML application experience, and understanding of electrical engineering fundamentals such as signal integrity, power delivery, and parasitic extraction.

Role involves developing open-source software to serve GPU-hosted LLMs (vLLM, SGLang, TensorRT-LLM), extend Dynamo for agentic inference (long-horizon reasoning, tool calling, multi-turn state), optimize inference-state management across memory/storage (KV/prefix cache, NIXL), and evolve distributed inference frontends. Requires MS/PhD or equivalent, 10+ years experience, and strong Rust and Python skills.

Senior security research role (12+ years) focused on validation and patching methods for defined vulnerability classes. Responsibilities include confirming exploitable issues, producing and verifying safe fixes, defining quality and revalidation standards, documenting evidence for independent review, and using AI-assisted workflows while maintaining rigorous security judgment.

NVIDIA is hiring a System Software Engineering Lead to work on a new Linux kernel-mode driver for NVIDIA GPUs, influence driver architecture, and upstream patches to the Linux kernel. The role requires 10+ years of software development experience and direct experience with device drivers and Linux kernel development.

NVIDIA is hiring a Senior Software Engineer to develop Rust APIs and libraries for CUDA Core Libraries. Responsibilities include building safe Rust abstractions over native CUDA and C/C++ interfaces, developing required C/C++ components, optimizing across Rust/C++/GPU boundaries, and owning features end-to-end (design, testing, profiling, packaging, release). The role requires extensive experience in Rust and C/C++, systems and GPU programming, FFI, performance optimization, and working in multi-language codebases.

Role involves designing and implementing Rust-native compiler backends, IR frameworks, and JIT/AOT toolchains to compile Rust to CUDA/PTX and GPU machine code. Requires deep Rust and compiler-internals expertise, experience with LLVM/MLIR/PTX, CUDA and GPU programming, and at least 5 years of relevant compiler or GPU code-generation experience.

The role involves analyzing compiler impacts and runtime behavior to improve HPC application performance, porting applications to different platforms and programming models, and working with compiler development and application engineering teams. Requires a BS/MS or equivalent in Computer Science or related field, 5+ years programming experience, strong skills with Fortran/C/C++, parallel programming (OpenACC, OpenMP, MPI, CUDA), performance analysis and tuning, and strong mathematical fundamentals.

Lead cross-stack projects to evaluate and integrate GCC/LLVM toolchains, optimize performance using techniques such as LTO/PGO/AutoFDO/BOLT, implement profiling and validation workflows (PMU, perf, flamegraphs), automate migrations and measurements, and produce reusable playbooks. Requires 8+ years of systems software engineering experience, strong C/C++ and Linux knowledge, compiler/build-system expertise, and experience with performance profiling and CI environments.

NVIDIA is hiring a Senior Software Engineer to build idiomatic Python APIs, language bindings, algorithms, and runtime infrastructure on top of native C/C++ for CUDA Core Libraries. Responsibilities include design, implementation, testing, profiling, benchmarking, packaging, and maintenance. Requires 8+ years of software development experience and strong production programming skills in both Python and C/C++. Location noted as US (remote; Santa Clara referenced). Salary ranges provided (USD, annual).

NVIDIA is hiring a Senior AI & Deep Learning Compiler Engineer to develop compiler optimization algorithms, define public APIs, and implement compiler techniques for AI workloads targeting NVIDIA GPUs. The role requires 3+ years of relevant experience, strong skills in CUDA, C/C++, Python, compiler technologies (e.g., MLIR, LLVM, XLA, Triton), and GPU performance analysis. Locations include US (Austin, Santa Clara) and remote options.

Work at the intersection of research, product, and engineering to bring agentic AI research into production: prototype and benchmark methods (LLMs, agent harnesses, multimodal models), design secure-by-default agent workflows, integrate inference and runtime infrastructure, and collaborate across teams. Requires 8+ years' experience and advanced degree or equivalent.

Role involves designing and evaluating routing policies for LLM traffic, building and running agentic benchmarks, producing calibration data and routing profiles, and contributing design docs and code to open-source. Collaborate across NVIDIA to integrate software with the accelerated serving stack. Requires a Bachelor's/Master's or equivalent, 8+ years industry experience in deep learning frameworks, experience with LLM evaluations/benchmarks, strong ML and CS fundamentals; GPU/CUDA and distributed systems experience are strong pluses.

Join the CUDA Core Libraries team to design and implement foundational CUDA C++ libraries, parallel algorithms, utilities, and runtime abstractions. Responsibilities include end-to-end feature ownership (design, implementation, testing, profiling, benchmarking, documentation, release, and maintenance), optimizing GPU algorithms, and enabling interoperability with Python and Rust. Requires a BS/MS/PhD or equivalent and 8+ years of relevant software development experience with strong production C/C++ skills and practical CUDA or parallel/heterogeneous programming experience.

Join the High Performance Computing Software development team to develop production-grade communication frameworks and applications for large supercomputers and datacenters. Responsibilities include designing and implementing protocols and algorithms, working on performance, scalability, and co-design with networking chip teams. Requirements include ~5 years programming in C/C++, 3 years in Linux, deep knowledge of InfiniBand and Ethernet, computer architecture and OS, performance optimization, and an MSc or equivalent. Posted by NVIDIA. Listed base salary ranges: 152,000–241,500 USD (Level 3) and 184,000–287,500 USD (Level 4).

Work on building high-performance storage paths, client libraries, and filesystem frameworks for AI workloads across object stores, file systems, and hybrid cloud. Responsibilities include improving performance, scalability, reliability, and observability for large-scale GPU-based training and inference, collaborating with platform and SRE teams, and using modern engineering practices. Requires a BS (or equivalent) and 5+ years of software engineering experience. Posted salary ranges: 152,000–241,500 USD (Level 3) and 184,000–287,500 USD (Level 4).

Technical focal point for GPU firmware and system software with CSP/hyperscale customers. Responsibilities include driving firmware workstreams (VBIOS, InfoROM, microcontroller firmware), orchestrating large-scale firmware updates and rollbacks, synthesizing customer requirements into NVIDIA's roadmap, and identifying cross-CSP operational patterns to improve tooling, documentation, and tests. Requires 15+ years of experience in GPU system software/firmware or accelerator platform engineering.

Technical lead working with CSP engineering teams to ensure rack-scale software/firmware can be deployed and operated at fleet scale. Responsibilities include driving architecture alignment (fabric management, link health, error handling, serviceability, multi-component firmware orchestration), capturing CSP feedback, improving tooling and test strategy, and making system-level software/firmware tradeoffs. Requires 15+ years in system software, platform firmware, or large distributed systems.

Lead development of scalable mapping and 3D reconstruction systems that transform crowdsourced camera and sensor data into high-quality maps for self-driving vehicles. Requires 15+ years experience, strong C++ production engineering skills, and expertise in 3D computer vision, multi-view geometry, pose estimation, sensor fusion, and large-scale sensor data processing.