Sr. Software Engineer - AI Triton Kernels

AMD
San Jose
Workplace: HybridFull timeUSD 127,400 - 218,400 annuallyFunction: Software EngineeringEducation: bachelorsSkills: ["Collaboration","Hands-on problem-solving","Curiosity","Quantitative analysis","Debugging"]

Develop and optimize state-of-the-art Triton/Gluon GPU kernels for large-scale LLM and multimodal workloads. Partner with research, compiler, and hardware teams to co-design high-performance solutions for AMD Instinct accelerators, targeting Triton AMD backend performance across ROCm and the LLVM AMDGPU stack. Own and productionize critical kernels in vLLM and SGL, perform deep profiling and microbenchmarking, and resolve end-to-end correctness and performance issues from PyTorch runtimes through compiler IR and GPU backends.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
AMD
AMD
2 days ago

Sr. Software Engineer - AI Triton Kernels

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 21 hours agoStatus: Live

Job Summary

Develop and optimize state-of-the-art Triton/Gluon GPU kernels for large-scale LLM and multimodal workloads. Partner with research, compiler, and hardware teams to co-design high-performance solutions for AMD Instinct accelerators, targeting Triton AMD backend performance across ROCm and the LLVM AMDGPU stack. Own and productionize critical kernels in vLLM and SGL, perform deep profiling and microbenchmarking, and resolve end-to-end correctness and performance issues from PyTorch runtimes through compiler IR and GPU backends.
Location: San Jose
Workplace: Hybrid
Employment Type: Full time
Job Function: Software Engineering
Seniority: Mid level

Key Responsibilities

  • •Design, research, implement, and optimize high-performance Triton kernels for matmul, attention (flash/paged/grouped-query), MoE, and fused transformer workloads.
  • •Productionize critical Triton/Gluon kernels in vLLM and SGL, ensuring correctness, scalability, and peak throughput.
  • •Collaborate with compiler engineers to develop and maintain the Triton AMD backend across ROCm and the LLVM AMDGPU stack.
  • •Drive deep kernel-level optimizations across the AMD memory hierarchy and wavefront execution (wave32/wave64), including MFMA utilization and occupancy tuning.
  • •Profile, microbenchmark, and debug performance/correctness issues end-to-end across PyTorch, vLLM/SGL runtimes, Triton IR/MLIR, ROCm runtime, and the LLVM backend.

Pay and Benefits

Salary: USD 127,400 - 218,400 annually

Key Requirements

  • •Deep expertise in SIMT programming, parallel algorithms, GPU architecture, and performance engineering.
  • •Strong hands-on experience developing GPU kernels for AI/ML workloads, including Triton-based matmul, attention, and fused transformer kernels.
  • •Comfort working end-to-end from vLLM/SGL down to ISA-level performance tuning and rigorous quantitative analysis.
  • •Meaningful contributions to open-source projects (e.g., Triton, Torch, vLLM, SGLang, IREE, MLIR, LLVM, or ROCm) with an upstream-first mindset.
  • •Bachelor’s or Master’s degree in Computer Science, Computer Engineering, Electrical Engineering, or equivalent practical experience.
Experience:AI/MLOpen source
Education:Bachelor's
Skills:CollaborationHands-on problem-solvingCuriosityQuantitative analysisDebugging
Languages:En-us
Tech Stack:TritonGluonPyTorchVLLMSGLangROCmLLVMAMDGPUMLIRIREETorchCUDAFlash attentionPaged attentionGrouped-query attentionMoEMatmulTransformer kernels

Eligibility

Work Authorization:Authorization required. Sponsorship not provided.

Company Brief

AMD
Designs and produces semiconductor products including CPUs, GPUs, and adaptive SoCs for consumer, enterprise, and embedded markets, competing across PCs, data centers, and gaming industries.
Industry: Electronics Manufacturing
Company Size: Enterprise (1,001+ employees)
Revenue: USD 1B+
Growth: Public Company
Valuation: Public Company (Market Cap in USD)
Funding: IPO / Publicly Listed
Headquarters: Santa Clara, United States
Founded: 1969
Glassdoor
Glassdoor: 3.9
WebsiteLinkedIn