Software Development Engineer, C++, Networking

What you'll do

• Define and drive the software architecture for scale-up switch fabrics connecting AMD Instinct GPUs.
• Lead switch software integration including link bring-up, topology discovery, routing, and congestion management.
• Architect and develop GPU firmware components for system management, power, security, and fabric link management.
• Lead development and optimization of Linux kernel drivers for memory management and GPU-to-GPU communication.
• Drive performance and enablement of the ROCm stack, including runtime and communication libraries like RCCL.
• Own new-platform bring-up from pre-silicon emulation through first silicon and production qualification.
• Lead cross-layer debug and root-cause analysis of complex issues spanning silicon, firmware, and drivers.
• Define performance and reliability goals for collective communication and multi-GPU workloads.

What we're looking for

• Bachelor’s or Master’s degree in Computer Science, Computer Engineering, Electrical Engineering, or a related discipline.
• PhD in a related field (preferred).
• Recognized subject matter expertise in GPU and networking system software from firmware and drivers through the user-space stack.
• Experience leading architecture, development, bring-up, and optimization for scale-up switch networking and GPU firmware.
• Experience developing Linux kernel drivers for GPUs, including memory management, virtualization, and GPU-to-GPU communication.
• Experience with switch software integration including SONiC, SAI, and link bring-up.
• Proficiency in C/C++ and Python (experience with Go or Rust is a plus) (preferred).
• Extensive experience in system software development for GPUs, accelerators, NICs, or network switches (preferred).