19 hours ago
Responsibilities
- Analyze AMD GPU hardware specifications and collaborate with hardware architects on current and future kernel designs.
- Develop support for new instruction-set architecture and hardware/software optimization features in the tensile-lite code generator.
- Innovate and implement algorithms for GEMM and GEMM + X, including attention workloads.
- Profile GEMM performance across the stack and identify root causes of bottlenecks.
- Partner with customers and internal stakeholders to understand workloads, reproduce performance issues, and deliver targeted optimizations.
Requirements
- Strong command of GPU computer architecture, including compute units, register files, cache/LDS hierarchies, memory bandwidth, matrix cores or WMMA-style instructions, and instruction scheduling.
- Deep expertise in GEMM algorithms and mapping them to GPU hardware through tiling, blocking, register/LDS allocation, wave scheduling, and memory hierarchy utilization.
- Strong software engineering fundamentals and a record of building production-quality, high-performance software.
- Experience optimizing GPU compute kernels, ideally GEMM and attention workloads.
- Proficiency in C++ and assembly-level GPU programming, with working knowledge of Python for tooling and automation.
- Understanding of parallel computing, memory hierarchies, and hardware-software performance tradeoffs.
- Ability to communicate clearly and work with hardware architects, customers, and cross-functional engineering teams.
- Direct experience with a GPU compiler backend.
- Experience with AMD CDNA or RDNA architectures, or comparable architectures such as NVIDIA Hopper or Blackwell.
- Contributions to industry standards, publications, or patents related to GPU compute or matrix-multiply optimization.
- BS, MS, or PhD in computer science, computer engineering, or a related field with deep relevant experience.
About AMD
We care deeply about transforming lives with AMD technology to enrich our industry, our communities, and the world. Our mission is to build great products that accelerate next-generation computing experiences – the building blocks for the data center, artificial intelligence, PCs, gaming and embedded. Underpinning our mission is the AMD culture. We push the limits of innovation to solve the world’s most important challenges. We strive for execution excellence while being direct, humble, collaborative, and inclusive of diverse perspectives. AMD together we advance_