Reference Architecture for Generative AI Compute and Storage with Lenovo ThinkSystem

Generative AI (GenAI) and its subset of large language models (LLMs) have exploded on the scene to offer capabilities never seen. From model-based creation of original art and music to the ability to distill actionable information from the massive content on the internet, these models have become powerful tools for discovery and innovation.

Lenovo’s initiatives on this front are designed to mitigate a company’s risks that they encountered when developing infrastructure supporting Generative AI. This document focuses on the role external storage plays in a comprehensive GenAI solution. The sections that follow address the challenges, provide an optimized reference architecture, and discuss the appropriate solution hardware.

1 Introduction
2 Analytics Discussion
3 Architectural Overview
4 Storage Platforms
5 Hardware and Network Topology
6 Other Considerations
7 Appendix: Lenovo Bill of materials
8 Appendix
Resources

To view the document, click the Download PDF button.

Related product families

Product families related to this document are the following:

• Artificial Intelligence