Senior Software Engineer - Triton Kernels

at  Advanced Micro Devices Inc

PREFERRED EXPERIENCE:

• Programming experience on GPUs - HIP, CUDA, OpenCL or Triton
• ML experience or knowledge in one or more of the following areas - Transformers, image models, recommendation systems
• Ability to work independently, define project goals and scope, and lead your own development effort
• Experience using debuggers
• Familiarity with PyTorch or JAX
• Knowledge of MLIR, LLVM and GPU assembly and GPU architecture is a plus
• Familiarity with models like LLama, Mixtral and Gemma is a plus

WHAT YOU DO AT AMD CHANGES EVERYTHING

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_
Responsibilities:
AMD is looking for an AI software development engineer to develop ML kernels in the Triton kernel language. We are looking for an engineer who is passionate about optimizing Machine Learning GPU kernels and improving the performance of key applications and benchmarks. What you do directly impacts the performance of AMD GPUs and enables us to become a competitive solution for generative AI. Become a part of our high-impact and incredibly talented Triton kernels team.

KEY RESPONSIBILITIES:

• Develop ML kernels for matrix multiplication, Flash Attention and other ML operators
• Benchmark, perform competitive analysis and optimize your kernels to improve performance
• Propose novel optimizations to the Triton compiler
• Collaborate with the GPU architecture team to improve future generations
• Apply knowledge of software engineering best practices