Tencent
Posted 2w ago

高级GPU推理加速工程师

Tencent
Beijing, Beijing, China
OnsiteFull Time
Responsibilities
  • optimizing inference
  • developing operators
  • upgrading framework
Requirements
  • Proven experience with Linux development
  • System programming, C++
  • GPU/CUDA optimization
  • Familiarity with TensorFlow or PyTorch, and strong problem-solving
  • >3 years relevant experience
Technical tools mentioned
LinuxC++CUDATensorFlowPyTorch

Job description

Responsibilities

1.负责营销业务中稀疏+稠密大模型的异构推理架构升级与性能优化,支持Dense Scaling 持续提升;

2.负责营销业务中GPU算子,计算图编译优化等业务开发,支撑业务迭代升级;

3.积极跟进GPU等异构硬件的业内前沿技术,持续完善异构推理加速框架。

Requirements

1.熟悉Linux开发环境,良好的系统编程、数据结构、算法基础、熟悉C++语言开发;

2.熟悉GPU优化手段,熟悉CUDA等异构硬件开发技术栈,具有搜广推/LLM大模型推理优化经历优先;

3.至少熟悉一种主流深度学习框架TensorFlow/Pytorch等,熟悉深度模型算法优先;

4.对追求极致高性能充满热情,具有优秀的分析问题和解决问题能力,良好的团队合作意识和沟通能力。

About Tencent

Provides integrated internet services, digital entertainment, and cloud technology.

Year founded
1998
Employees
112771
Organization type
Public
Headquarters
CN