Principal Distributed Systems Software Engineer, Collectives

ADVANCE YOUR CAREER. ADVANCE THE WORLD. At AMD, we believe technology has the power to solve the world’s most important challenges. From advancing healthcare and scientific discovery to powering AI and the technologies people rely on every day, innovation at AMD is shaping the future. Whether you’re designing next-gen processors, enabling AI breakthroughs, or bringing leading edge products to market, every role at AMD contributes to something bigger — technology that moves the world forward.

Join us and, together, we’ll advance your career.

THE ROLE:

As a core member of the Collectives and Network Performance team, you will play a pivotal role in defining and delivering industry-leading AI pre-training and distributed inference performance on AMD GPUs. Your expertise will be critical in optimizing collective communication libraries, scale-up and scale-out networking software, and training/inference performance across multi-GPU and multi-node systems. You will engage with internal hardware, GPU library, compiler, runtime and ROCm teams as well as open-source maintainers to ensure seamless integration of optimizations, applying advanced engineering principles to drive continuous improvement across multiple generations of AMD GPUs and a wide range of state-of-the-art AI models.

THE PERSON:
Seeking an industry-leading expert in collective communication and network performance, with advanced technical and analytical skills in Linux environments and deep C++ expertise. The ideal candidate will excel in providing technical leadership, guiding teams, and driving projects and initiatives independently. You will define goals, scope, and own development efforts while collaborating effectively within a high-performing team. You have deep knowledge of network, NIC and GPU hardware architecture, hands-on experience mapping model architectures to low-level software and hardware, and a strong grasp of current trends in distributed training and inference  RESPONSIBILITIES:
Drive Collectives & Network Strategy:
Define goals, scope and roadmap for a major area of AMD’s collectives and network optimizations, and align them with hardware, framework and ROCm roadmaps.

Architect Collective Communication Libraries:
Lead the design and development of collective algorithms and optimizations (e.g., all-reduce, all-to-all, all-gather) across scale-up and scale-out networks.

Optimize Distributed Training & Inference:
Lead performance tuning, profiling and analysis of large-scale LLM, diffusion, multimodal, Rec Sys and generative AI models on single-node and multi-node systems, and drive the tradeoffs and design decisions.

Hardware-Software Co-Design:
Participate in co-design of future scale-up networks, NICs and scale-out networks, and represent software performance requirements to hardware teams.

Performance Modeling & Tooling:
Drive frameworks, tools and infrastructure for performance estimation, modeling, regression tracking and reporting.

Collaborate with Framework & Library Teams:
Work closely with internal GPU library, compiler and runtime teams, and with open-source maintainers (e.g., PyTorch, JAX, vLLM, SGLang), to ensure optimizations are aligned with requirements and integrated upstream.

Customer & Partner Guidance:
Provide guidelines on efficient network load balancing, workload scheduling and model sharding strategies.

Communicate with Leadership:
Present performance analysis and modeling results to stakeholders and senior leadership, with concrete recommendations.

Software Engineering Best Practices:
Apply sound engineering principles and set standards to ensure robust, maintainable solutions.

Lead, Guide & Mentor:
Provide strategic direction and mentorship to junior team members, fostering growth and…