AMD is seeking a Fellow Software Development Engineer to develop features in the tensile-lite assembly code generator for General Matrix Multiplication (GEMM) and GEMM + X. This role focuses on optimizing AMD's GPU architecture for AI workload efficiency and requires deep hardware-aware optimization skills. The engineer will work closely with hardware architects and innovate algorithms to enhance performance. This position is ideal for someone with a strong background in GPU microarchitecture and low-level code generation.

• Strong command of GPU computer architecture, including compute units and memory bandwidth.
• Deep expertise in GEMM algorithms and their mapping onto GPU hardware.
• Strong software engineering fundamentals with a track record of high-performance software.
• Demonstrated experience optimizing GPU compute kernels, particularly GEMM and attention.
• Proficiency in C++ and assembly-level GPU programming, with knowledge of Python for tooling.

• Analyze AMD GPU hardware specifications and align kernel design with silicon capabilities.
• Develop support for new ISA and HW/SW optimization features in the code-generator.
• Innovate and implement new algorithms for GEMM and GEMM + X.
• Profile and root-cause GEMM performance bottlenecks across the stack.
• Partner with customers and stakeholders to understand workload requirements and deliver targeted performance.

• AMD provides a competitive 'Total Rewards' package that focuses on financial growth, health, and work-life balance.