AI Framework and Performance Experts(vLLM, SGLang)
Core
Optimize and develop deep learning frameworks (TensorFlow, PyTorch, SGLang) and GPU kernels for AMD GPUs to accelerate training and inference at scale.
Role type
Senior IC GPU software engineer (AI frameworks & kernels)
Builds
High-performance GPU kernels, optimized deep learning frameworks, and distributed inference/training systems
Domain
AI/ML infrastructure, GPU computing, compiler technology
Deliverable
production ML models
Required skills
C++, Python, GPU kernel development, deep learning framework optimization, distributed systems, graph compilers, open-source collaboration
Preferred skills
SGLang, LLM optimization, compiler design (LLVM, MLIR), ROCm, HIP/CUDA, GCN/CDNA architecture, heterogeneous cluster workloads
Technologies
HIP, Triton, ROCm, XLA, TorchDynamo, TensorFlow, PyTorch, SGLang, LLVM, MLIR
Responsibilities
Optimize deep learning frameworks via upstream contributions; Develop and tune large-scale training/inference models; Design and optimize GPU kernels; Collaborate with compiler and GPU library teams; Contribute to SGLang development; Tune distributed multi-GPU/multi-node systems; Integrate graph compilers; Partner with open-source maintainers
Seniority
Senior, hands-on IC
