System Software Engineer – Data Center GPU Compute Diagnostics
Core
Develop low-level diagnostic and stress software for next-generation data center GPUs and rack-scale AI systems to validate hardware, debug failures, and bring up new silicon.
Role type
System Software Engineer (Hardware Diagnostics & Validation)
Builds
Diagnostic and stress software for GPU compute engines, memory subsystems, and high-speed interfaces.
Domain
Data Center AI / GPU Hardware Validation
Deliverable
production ML models
Required skills
C/C++, Python, Embedded/Firmware development, Linux device drivers, Hardware debugging, Computer architecture (memory, caches, DMA, buses), Failure analysis
Preferred skills
CUDA programming, GEMM-style compute, NCCL, PyTorch-based workloads, Silicon bring-up experience
Technologies
C, C++, Python, Linux, CUDA, NCCL, PyTorch
Responsibilities
Develop diagnostic and stress software in C/C++ and Python; Interface with hardware blocks, firmware, Linux device drivers, and registers; Bring up and validate new silicon and system features; Create targeted tests for compute engines, memory, and interfaces; Investigate hardware and software failures; Use modern development and analysis tools.
Seniority
Mid-level, hands-on IC
