Server Performance Architect - Hardware
Core
Drive architectural performance for next-generation AI server systems by bridging deep architectural knowledge, workload analysis, and hands-on silicon investigations.
Role type
Senior IC server performance architect (hardware)
Builds
Next-generation AI server systems for data centers
Domain
Hardware architecture, AI/ML infrastructure, High-Performance Computing (HPC)
Deliverable
production ML models
Required skills
Server/system performance architecture, CPU/GPU/memory subsystem knowledge, System-level profiling and analysis, Python/C++ scripting, Trade-off studies, Cross-functional collaboration
Preferred skills
AI/ML training and inference workloads, NVIDIA GPU architectures (Hopper, Blackwell), Chip-to-chip interconnect analysis, Power/thermal-aware optimization
Technologies
CUDA, NCCL, PCIe/CXL, DDR/HBM, C2C, UCIe
Responsibilities
Define server-level performance targets across subsystems, Conduct hands-on workload characterization and bottleneck analysis, Root-cause performance issues using profiling tools, Perform trade-off studies on topology and thermal/power envelopes, Collaborate with silicon/firmware/software teams, Develop automation for performance regression tracking, Represent performance perspective in architecture reviews, Build analytical performance models and simulation frameworks
Seniority
Senior, hands-on IC
