Intern, AI Systems Performance Engineer

Profile AI agent workloads and memory hierarchies across host DDR5, GPU HBM, and Samsung CXL memory expanders.

Originally posted 13 November 2026 by the employer.

About the role

This role researches memory solutions for AI workloads, focusing on HBM, DRAM, and CXL memory expansion. You will benchmark and profile how agentic AI workloads stress server memory subsystems using Samsung CXL memory expanders and NVIDIA GPUs.

What you'll do

• Build runtime adapters and a trace record/replay harness for AI agent frameworks.
• Develop micro-kernel and end-to-end agent workload benchmarks that stress CPU, memory, storage, and accelerator subsystems.
• Profile agent workloads with perf, eBPF, hardware PMUs, and NVML/DCGM to break down per-turn latency and memory bandwidth usage.
• Measure host DDR5, GPU HBM, and CXL.mem bandwidth and latency across Intel Xeon and AMD EPYC servers.
• Evaluate multi-agent KV-cache sharing and CXL memory expansion on Samsung CXL expanders.
• Build an initial memory tiering cost model (HBM/DDR5/CXL) for long-running MoE agent workloads.

What you'll need

• Currently pursuing a PhD in Computer Science, Computer Engineering, Electrical Engineering, or a related field.
• Programming skills in Python and C/C++ in a Linux environment.
• Understanding of computer architecture, including CPU caches, the memory hierarchy, and NUMA.
• Experience with performance profiling tools such as perf, eBPF, VTune, or Nsight.

Nice to have

• Hands-on experience with LLM serving frameworks (vLLM, SGLang) or GPU programming.
• Familiarity with HBM, CXL, or heterogeneous memory systems.
• Experience with AI agent frameworks or benchmark development.

Skills: CXL, HBM, DDR5, perf, eBPF, vLLM

More semiconductor jobs: Samsung Semiconductor jobs · Hardware design jobs · San Jose jobs