Systems Performance Engineer, AI Solutions Lab
Core
Drive end-to-end performance optimization for AI serving and training stacks, identifying bottlenecks across the full data path from application to infrastructure layers.
Role type
Senior IC systems performance engineer (AI infrastructure)
Builds
Validated AI platform performance proof points and live demos for hyperscalers and customers
Domain
Semiconductor / AI Infrastructure / High-Performance Computing
Deliverable
production ML models
Required skills
AI serving frameworks (vLLM, SGLang, TensorRT-LLM), distributed training frameworks (PyTorch DDP/FSDP, DeepSpeed, Megatron-LM), GPU profiling tools (Nsight, PyTorch Profiler, ROCm), RCCL/NCCL, RDMA, Python, Linux, containerized environments (Docker, Kubernetes)
Preferred skills
Go, GPU kernel analysis, system tracing, serving-stack internals (continuous batching, paged KV-cache, quantization), customer PoC experience
Technologies
vLLM, SGLang, TensorRT-LLM, PyTorch, DeepSpeed, Megatron-LM, Nsight Systems, Nsight Compute, ROCm, Docker, Kubernetes, RCCL, NCCL, RDMA
Responsibilities
Run AI benchmarking experiments mirroring real customer workloads; operate GPU/HPC lab clusters; validate Teralynx switching and AI fabrics; profile serving-stack internals (quantization, scheduling, KV-cache); analyze results to guide kernel and system-level optimizations; create performance summaries and artifacts for customer presentations and executive reviews.
Seniority
Senior, hands-on IC
