AMD is seeking a Senior Software Development Engineer to focus on high-performance collective communication libraries and GPU communication runtimes for AI and HPC workloads. The role involves optimizing software for multi-GPU and multi-node communication across AMD's ROCm software stack. The engineer will collaborate with various teams to validate interconnect performance and resolve complex issues. This position is integral to advancing AMD's technology and supporting the development of next-generation accelerator platforms.

• Master's degree or foreign equivalent in Computer Science, Computer Engineering, Electrical Engineering, or a related field.
• Experience in Python, C, or C++.
• Experience in developing benchmarks to test software performance.
• Experience in GPU software programming and development.
• Experience in software debugging on macOS, Linux, or Unix.
• Knowledge of data structures and multi-threading.
• Familiarity with CUDA.

• Research, design, develop, and test high-performance collective communication libraries.
• Optimize, troubleshoot, and debug software for efficient multi-GPU and multi-node communication.
• Contribute to open-source libraries for GPU communication and benchmarking tools.
• Develop and tune collective algorithms for distributed training and inference frameworks.
• Design micro-benchmarks and performance characterization tools to evaluate transfer performance.
• Collaborate with hardware, firmware, driver, and AI framework teams to enable new silicon.
• Analyze complex performance data and architectural factors to influence product decisions.

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