Software Engineer

NVIDIA · Taipei, Taiwan We are seeking a motivated software engineer to develop diagnostics and validation software for NVIDIA's next-generation GPU servers. You will develop production C++ software on Linux, investigate challenging system failures, and collaborate with software, firmware, and hardware teams. Responsibilities Develop and maintain MODS diagnostic software, stress workloads, and validation tools for GPU boards, servers, and racks. Implement tests covering GPU compute, memory, power, thermal behavior, PCIe, NVLink, and other platform components, with clear pass/fail criteria and useful failure information. Write reliable, efficient C++ code and improve the performance, maintainability, and test coverage of a large production codebase. Reproduce and debug complex failures across software, operating systems, firmware, and hardware, working with the relevant teams to identify root causes and verify fixes. Integrate diagnostics into manufacturing and data center workflows, balancing test coverage, runtime, and reliability. Develop automation and supporting tools using Python and JavaScript; improve regression testing, log analysis, and diagnostic reporting. Apply AI-assisted tools to code development, debugging, test generation, and data analysis, reviewing and validating their results. Collaborate with architecture, driver, firmware, hardware, manufacturing, and partner teams, and document test behavior and troubleshooting procedures. Requirements BS or MS in Computer Science, Computer Engineering, Electrical Engineering, or a related field, or equivalent practical experience. More than 3 years of professional software engineering experience, with substantial hands-on C++ development for production environments. Strong C++ programming skills, including sound understanding of data structures, algorithms, memory management, and maintainable software design. Extensive debugging experience: reproducing difficult failures, forming and testing hypotheses, analyzing logs and program state, and verifying root-cause fixes. Strong Linux development and troubleshooting skills, with a solid foundation in operating systems, processes, threads, memory, and concurrency. Familiarity with Python and JavaScript for automation, tooling, or supporting applications. Nice to Have Experience with GPU or server diagnostics, system validation, manufacturing tests, or field-failure analysis. Familiarity with PCIe, NVLink, InfiniBand, Ethernet, or x86/ARM server platforms. Experience with CUDA, OpenGL, drivers, or software that interacts closely with hardware. Practical use of GDB, sanitizers, profiling tools, or tracing tools to diagnose correctness and performance issues. Experience improving automated regression pipelines, diagnostic coverage, or structured logging. Benefits Competitive salaries and a generous benefits package.