20 hours ago
Santa Clara, CA, USAStaff+
Responsibilities
- Define and drive software architecture for scale-up switch fabrics connecting AMD Instinct GPUs, including Ethernet, UALink, and Ultra Ethernet.
- Lead switch software integration, link bring-up, SerDes and retimer tuning, FEC/BER characterization, routing, congestion management, topology discovery, and telemetry.
- Architect and develop GPU firmware for system management, power and thermal management, security, RAS, and fabric/link management, including PLDM, Redfish, and BMC integration.
- Lead Linux kernel driver development and optimization for Instinct GPUs, including memory management, GPU-to-GPU communication, partitioning, virtualization, error handling, and recovery.
- Drive ROCm enablement and performance, including runtime and RCCL integration with PyTorch, JAX, vLLM, and SGLang.
- Own platform bring-up from emulation and simulation through first silicon, system integration, and production qualification.
- Lead cross-layer debugging, root-cause analysis, and optimization across silicon, firmware, drivers, fabrics, and applications.
- Set technical direction, mentor engineers, and coordinate execution across globally distributed teams and external partners.
Requirements
- Bachelor’s or master’s degree in Computer Science, Computer Engineering, Electrical Engineering, or a related discipline; PhD is a plus.
- Extensive system software development experience for GPUs, accelerators, NICs, or network switches.
- Strong data-center networking background, including Ethernet, high-speed SerDes, FEC, RDMA/RoCE, congestion control, and switch ASIC programming.
- Hands-on firmware development experience for SoCs or accelerators, including embedded C/C++, boot flows, power management, security, and BMC or out-of-band management.
- Strong Linux kernel and device-driver development experience, including PCIe, DMA, IOMMU, memory management, and interrupt handling.
- Knowledge of GPU architecture, ROCm/HIP, CUDA or similar GPU stacks, and collective communication libraries such as RCCL/NCCL and MPI.
- Familiarity with PyTorch, JAX, TensorFlow, and distributed training or inference at scale.
- Demonstrated experience leading hardware-platform bring-up and debugging hardware/software interactions with JTAG, protocol analyzers, and logic analyzers.
- Proficiency in C/C++ and Python; Go or Rust experience is a plus.
- Ability to set technical direction and influence multiple teams and organizations, with strong communication, collaboration, and leadership skills.
Benefits
- AMD benefits are described in the AMD benefits at a glance materials.
- This role is not eligible for visa sponsorship.
About AMD
AMD designs and sells CPUs, GPUs, and adaptive/embedded computing products for PCs, data centers, gaming, and edge devices. Its portfolio includes Ryzen and EPYC processors, Radeon and Instinct graphics, and adaptive SoCs from its Xilinx acquisition, sold to OEMs, cloud providers, and device makers. Founded in 1969 and headquartered in Santa Clara, it is a public company on NASDAQ and supplies semi-custom chips for major game consoles.