高性能计算专家（北京/深圳/上海）
Core
Build industry-leading high-performance computing platforms by optimizing underlying compute performance across NVIDIA, Ascend, Intel, and AMD hardware.
Role type
Senior IC high-performance computing expert (GPU/NPU)
Builds
High-performance compute platforms for training and inference workloads
Domain
AI infrastructure / High-performance computing
Deliverable
production ML models
Required skills
GPU hardware architecture expertise, CUDA/ROCm programming, multi-GPU/NPU optimization, deep learning framework optimization, compute diagnostics and tuning, kernel virtualization, academic research
Preferred skills
Management experience, industry foresight
Technologies
NVIDIA CUDA, AMD ROCm, PyTorch, TensorFlow, NVIDIA GPUs, Ascend NPUs, Intel/AMD CPUs
Responsibilities
Optimize underlying compute performance and build high-performance platforms; Troubleshoot hardware, driver, and compute library issues for GPUs/NPUs; Conduct academic research, publish papers, and present at conferences; Evaluate and integrate emerging technologies into existing systems.
