NVIDIA's Performance Lab (PerfLab) builds the systems and automation used to evaluate the performance and quality of accelerated computing and AI workloads. We turn complex benchmark experiments into reliable, scalable, and reproducible workflows that help engineering teams make better decisions faster.
We are looking for an experienced and highly self-motivated System Software Engineer to help build the next generation of PerfLab's benchmark infrastructure. You will independently own meaningful platform components and take projects from problem discovery and technical design through production deployment and adoption. The ideal candidate enjoys finding important engineering problems, understanding their root causes, and using technology to create simple, reusable solutions.
You will collaborate with NVIDIA teams around the world and work on evolving areas such as large language models, agentic AI, accelerated computing, and other emerging AI workloads.
What You'll Be Doing:
Design, build, and maintain reusable software, services, and workflows that automate benchmark definition, execution, result collection, validation, and reporting across local, cluster, and cloud-native environments.
Carry out performance testing and analysis as needed. Develop a deep understanding of existing performance workflows and infrastructure, and translate real-world needs into scalable, user-friendly solutions.
Improve the reliability, scalability, observability, and reproducibility of large benchmark campaigns.
Diagnose complex issues across applications, Linux systems, containers, distributed jobs, compute resources, networking, and storage.
Build strong partnerships with performance engineers, QA teams, product teams, and other customers; agree on goals, organize execution, and drive new ideas and projects from concept through adoption.
Contribute to technical designs, code reviews, documentation, and internal or open-source infrastructure projects.
Apply AI-assisted automation where it can meaningfully improve benchmark creation, failure triage, data analysis, or engineering productivity.
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 in practice.
5+ years of relevant software engineering experience in system software, infrastructure, developer platforms, distributed systems, or production automation.
Strong Python programming and software engineering skills, with experience building production-quality tools, services, or automation frameworks.
Solid understanding of Linux and system-level concepts such as processes, concurrency, networking, storage, resource management, and failure handling.
Hands-on experience with containers and at least one workload orchestration, scheduling, or distributed computing platform, as well as designing reliable systems or pipelines with clear interfaces, testing, observability, and recovery behavior.
Working knowledge of machine learning, AI, or accelerated-computing workloads and an interest in how their performance and quality are evaluated.
Strong analytical and problem-solving abilities, with the capacity to manage multiple priorities effectively and adapt in a dynamic, fast-changing environment.
High self-motivation and demonstrated ability to identify valuable problems, turn ambiguous needs into clear technical plans, and drive projects through delivery and adoption.
Excellent communication and organizational skills, with the ability to align cross-functional stakeholders, collaborate with globally distributed teams, and move new ideas toward concrete outcomes.
Ways to Stand Out From the Crowd:
Experience with GPU or AI infrastructure, distributed training or inference, model evaluation, or performance benchmarking.
Experience building workflow engines, schedulers, experiment platforms, test frameworks, or developer infrastructure.
Experience operating distributed or cloud-native systems in production, including performance profiling, capacity analysis, resource scheduling, or multi-node workloads.
Practical experience using AI agents, tool-calling, or coding agents to automate engineering workflows.
Contributions to open-source infrastructure or developer tooling, or a track record of initiating automation that reduced manual work, improved reliability, or helped other teams move faster.
We have some of the most forward-thinking and hardworking people in the world working for us. If you are creative, autonomous, and passionate about building systems that make complex AI performance work repeatable and scalable, we want to hear from you.
