模型推理优化工程师（J106570）
Core
Optimize inference performance for generative AI models (LLMs, Diffusion, video generation, digital humans) to increase throughput, reduce latency, and lower compute costs.
Role type
Senior IC machine-learning inference optimization engineer
Builds
High-performance inference systems for generative AI products
Domain
Generative AI, Deep Learning, High-Performance Computing
Deliverable
production ML models
Required skills
C/C++, Python, CUDA programming, TensorRT, ONNX Runtime, vLLM, model quantization, operator fusion, memory management, profiling tools
Preferred skills
Triton kernel development, heterogeneous computing (GPU/CPU), model distillation and pruning
Technologies
TensorRT, ONNX Runtime, PyTorch, vLLM, TensorRT-LLM, CUDA, Triton
Responsibilities
Optimize inference frameworks and operators for specific model architectures; implement model compression techniques; design heterogeneous computing solutions; profile and resolve performance bottlenecks in compute and memory access.
Seniority
Senior, hands-on IC
