CareerPlanSign in

高性能网络技术方案架构师

Shenzhen, China💼 Full-time🗓 2026-09-28

Core

Design and implement high-performance network solutions for AI training and inference workloads, optimizing communication between heterogeneous GPU clusters.

Role type

Senior IC high-performance network solutions architect

Builds

High-performance network infrastructure for AI training/inference frameworks

Domain

AI infrastructure / High-performance computing

Deliverable

production ML models

Required skills

RDMA (InfiniBand/RoCE), SmartNIC, DPU, PyTorch, TensorFlow, NCCL, DeepEP, distributed training optimization, heterogeneous GPU deployment

Preferred skills

Heterogeneous GPU platform adaptation (NVIDIA/Ascend/AMD), data center network architecture design, network productization

Technologies

InfiniBand, RoCE, NVIDIA, Ascend, AMD, PyTorch, TensorFlow, NCCL, DeepEP

Responsibilities

Collaborate with business teams to decompose network product requirements and drive implementation; Design technical solutions and conduct PoC validation for key network technologies; Introduce and optimize heterogeneous GPU card deployments for training/inference frameworks; Explore new high-performance network directions and assess technical feasibility and business value.

Seniority

Senior, hands-on IC

Sourced via tencent · Listed on CareerPlan, which tracks 853,000+ jobs from 20+ sources.