GPU服务器硬件架构及性能优化工程师-服务器平台
Core
Define GPU hardware product roadmaps, validate performance across scenarios, and optimize system bottlenecks for cloud AI accelerators.
Role type
Senior IC GPU server hardware architecture and performance optimization engineer
Builds
Self-developed hardware server products and optimized GPU solutions for cloud AI workloads
Domain
Cloud computing, AI infrastructure, heterogeneous computing
Deliverable
production ML models
Required skills
GPU architecture principles, DPDK, Linux kernel internals, virtualization, performance profiling tools (Speccpu, Fio, Iperf, Stream, MLperf), hardware component selection (CPU, storage, NIC, SAS)
Preferred skills
Internet business component performance testing, stress/load testing, cross-platform scenario validation
Technologies
Linux, DPDK, GPU drivers, SAS, NIC, Fio, Iperf, Stream, MLperf
Responsibilities
Design and execute performance testing schemes to analyze system bottlenecks; Evaluate business scenario benefits to determine selection strategies; Monitor and analyze quality/performance of heterogeneous cloud hardware in production; Coordinate technical problem resolution during new hardware/technology deployment; Output self-developed hardware server product documentation.
Seniority
Senior, hands-on IC
