Live opening · Posting date unavailable
Tencent · Shanghai, China
At a glance
The key details from the original listing.
CompanyTencent
LocationShanghai, China
Experience> 5 years of working experience
Source
Own
ListedPosting date unavailable
About the role
Description supplied by the original job listing.
1.超大规模LLM性能工程: 主导并规划千亿参数级大模型的极致性能优化技术路线。负责 vLLM/TensorRT-LLM 等主流推理框架的内核级优化与落地;
2.低比特与稀疏模型优化: 牵头 INT4/FP8/AWQ 等前沿低比特量化技术的工业级系统化落地,平衡精度与计算效率。并设计面向 MoE 模型的分布式调度、路由、显存管理及跨卡通信的优化方案;
3.统一与多模态架构: 定义并设计一套具备长期扩展性的统一 AI 推理引擎架构,以支撑自回归生成任务,并前瞻性地解决多模态大模型(如视觉-语言模型)的协同推理部署挑战;
4.异构算力与适配: 主导推理引擎在AI芯片(如昇腾、海光、天数等)平台上的战略级移植、生态适配与性能优化。对 HCCL/NCCL 等通信原语进行深度优化和定制,实现跨异构架构的算力自主可控;
5.核心算子优化与指令架构创新 (Enhanced Focus):深度介入 GPU/NPU 硬件底层,主导设计和实现LLM特有高性能算子;
6.负责海外机房建设进展跟综及GPU、网络等相关验收。
Experience
> 5 years of working experience
More openings worth a look
Recently tracked roles with full details and direct application links.
Good roles move before most people even see them.
Tell JobBeeper what you want and get fresh matches delivered in minutes.
Start your free trial →