Intern, AI System Architecture Research Scientist

Samsung Semiconductor•San Jose, CA
•$28 - $68•Onsite

About The Position

The Architecture Research Lab is researching AI system architectures to address memory bandwidth and communication bottlenecks. This internship focuses on how memory systems should evolve for agentic AI, whose long-horizon, multi-turn workloads accumulate context faster than memory capacity and bandwidth grow. The intern will research memory architectures and system optimizations to maintain accurate, efficient inference; evaluate them using performance models and accuracy studies; and translate research findings into publications and requirements for future memory products. The work spans AI system architecture up to the rack level, including memory hierarchies that combine HBM, near-memory compute, High Bandwidth Flash, and host or CXL memory. The intern will characterize LLM and agentic workloads, study placement and movement of model weights and KV cache, develop performance models and simulators, explore architectural tradeoffs, and assess accuracy impacts using open models and relevant benchmarks. The role collaborates with architects, memory technologists, and system-software researchers, and contributes to publications, invention disclosures, and future product directions. The position requires daily onsite presence at Samsung’s San Jose headquarters and reports to a Principal Engineer.

Requirements

  • Currently pursuing a PhD and have at least one academic quarter or semester remaining.
  • Strong background in AI system architecture and memory systems, including memory hierarchy, DRAM/HBM, and data movement.
  • Deep understanding of LLM and agentic inference, including prefill and decode, KV cache, and long-context workloads, and how memory optimizations such as KV-cache compression or eviction affect model accuracy.
  • Experience in performance analysis and modeling of AI systems, including roofline analysis and building or extending simulators or analytical models to evaluate architecture ideas.
  • Hands-on research experience managing model state across memory tiers, such as KV-cache paging, reuse, offloading, compression or eviction, or weight offloading for large and MoE models, using serving stacks such as vLLM, SGLang, or LMCache.
  • Strong programming skills in Python and C/C++, and experience with PyTorch.
  • Daily onsite presence at the San Jose headquarters.
  • Collaborative, innovative, and creative approach, including building relationships, welcoming new approaches, exploring ideas, and adapting to change.
  • Curiosity and resilience, with a willingness to seek data to build understanding.
  • Inclusive approach and ability to adapt communication and working style to diverse global norms.

Nice To Haves

  • Three or more years studying Computer Science.
  • First author or major contributor to publications at top-tier computer architecture, systems, or machine learning venues.

Responsibilities

  • Characterize LLM and agentic inference workloads, including multi-turn sessions, long and growing context, and reuse across requests, and derive memory capacity, bandwidth, and latency requirements.
  • Research AI inference memory hierarchies combining HBM, near-memory compute, emerging tiers such as High Bandwidth Flash, and host or CXL memory.
  • Study how model weights and KV cache are placed, moved, reused, and compressed across memory tiers.
  • Propose architectures and system optimizations, spanning hardware features, runtime, and memory-management policies, to maintain long-horizon agent accuracy within memory limits.
  • Build and extend performance models and simulators, and conduct design-space explorations across throughput, capacity, and cost metrics.
  • Measure the accuracy impact of memory optimizations on long-context and agentic benchmarks with open models.
  • Drive research from problem definition through results, collaborating with architects, memory technologists, and system-software researchers.
  • Publish results at top-tier venues, contribute invention disclosures, and help turn research into product directions.
  • Complete other responsibilities as assigned.

Benefits

  • Medical, dental, and vision benefits.
  • Relocation stipend for employees living more than 50 miles from the new office.
  • Monthly housing stipend for eligible summer interns signing a new lease.
  • Charitable giving match and opportunities to get involved in the community.
  • On-demand emotional wellness apps and paid therapy sessions.
  • Onsite café and gym, plus virtual fitness classes.
  • Flexible environment to help employees find work-life balance.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service