← All roles

Senior GPU Compiler Development Engineer

AMD· Engineering
Apply on company site
Beijing, ChinaFull-timePosted 3mo ago

About the role

AMD is seeking an MLIR compiler engineer to represent and lower deep-learning operators for next-generation models on AMD GPUs. You will help architect dynamic-graph IR lowering, implement operator lowerings, and contribute to MLIR dialect conventions, working on a single dynamic-shape graph.

Responsibilities

  • Help architect the dynamic-graph IR lowering on a single graph
  • Implement operator definitions and MLIR lowerings for attention and control-flow operators required by new models
  • Build model- and operator-level runtime-status reporting and analyze graph-optimizer behavior
  • Contribute to MLIR dialect conventions and the operator-add checklist; review operator-lowering work
  • Stay informed of compiler, algorithm, and architecture trends and apply them to the IR stack
  • Debug and fix existing issues and research alternative, more efficient ways to accomplish the same work
  • Develop technical relationships with peers and partners

Qualifications

  • Strong object-oriented programming background, C/C++ preferred, with a keen attention to detail
  • Hands-on experience with MLIR (dialects, lowering, passes)
  • Solid understanding of deep-learning operators and computational graphs
  • LLVM, GPU programming (HIP / CUDA), and ONNX / ONNX Runtime experience is a plus
  • Attention / transformer and diffusion operator internals; dynamic-shape and quantized inference is a plus
  • Experience with Windows and Linux development; familiarity with debuggers, source code control systems (GitHub), and profilers
  • Regular, hands-on use of AI coding agents / assistants to improve day-to-day engineering efficiency
  • Professional working proficiency in English for cross-site collaboration
  • Effective communication and problem-solving skills

Skills

Required:C/C++ programmingMLIRDeep-learning operators and computational graphsEnglish communication
Nice to have:LLVMGPU programming (HIP / CUDA)ONNX / ONNX RuntimeWindows & Linux development

Benefits

  • AMD benefits at a glance
Last updated Oct 22, 2026. Interested? Apply on the company site.
AMD logo

About AMD

together we advance_AI

AMD delivers leadership high-performance and adaptive computing solutions to advance data center AI, AI PCs, intelligent edge devices, gaming, & beyond.

AIData CenterHigh-Performance ComputingGamingEdge Devices
Company profileWebsite

More open roles at AMD