Systems Software Engineer, (Linux, CUDA)

Overview Join Micron’s Storage and Memory Systems group to drive system-level innovations that unlock the value of next-gen hardware. You will work with AI/ML experts, hardware engineers, and system architects to prove concepts and quantify the impact of memory and storage technologies. Expect hands-on software development, performance analysis, and cross-functional collaboration to shape future Micron products. You’ll use emulators and heterogeneous compute skills to forecast system behavior before hardware is available. This role offers a meaningful path to influence memory system design in a fast-moving, pathfinding environment. Compensation / Benefitsmedical, dental and vision planspaid time-offpaid holidaysincome protection programspaid family leavebenefits guide on micron.com/careers/benefits ResponsibilitiesDevelop software and port applications to demonstrate and measure the value of proof-of-concept hardware systems and featuresCollaborate cross-functionally with AI/ML experts, software leaders, hardware engineers, and system architects to identify memory- and storage-related opportunities for future AI challengesDefine, set up, and implement analytical experimentsBuild and modify Linux applications and libraries using C, C++, Python, Rust, SYCL, and/or CUDA for advanced heterogeneous systemsDevelop and use emulators to predict system-level impacts of proposed capabilities before hardware availabilityEmbrace emerging AI tools to accelerate productivityPresent innovations and insights to internal customers across Micron Key requirementsBachelor’s (or higher) in CS, CE or related field with substantial experienceExperience with GPU programming (CUDA, SYCL, OpenCL, PyTorch) and knowledge of GPU memory hierarchiesUnderstanding of modern AI pipeline stages and features (transformers, neural networks, graphs, tokenization)Strong C, C++, Python, and shell scripting skills; Linux programming expertise (multithreading, IPC, ioctl, gdb, /proc, /sysfs)Experience developing software to enable or demonstrate hardware system capabilitiesDeep understanding of heterogeneous compute features (NUMA, caching, GPUs, PCIe, CXL, NVLink, UALink)Proven ability to debug system‑level issues across app, kernel, libraries, OS config, and hardware boundariescross-functional collaborationproblem solvingcommunicationCUDASYCLOpenCL