Top
Authors
- Hapsara Sukasdadi
- Mike O'Brien
- Devin Bost
- Ka Leung
Published
7 Oct 2026Form Number
LP2550PDF size
50 pages, 3.2 MBSubscribed to LP2550.
Thank you for your feedback.
Abstract
Lenovo Hybrid AI 281 with IBM is a compact, single-node enterprise AI platform designed for retrieval-augmented generation (RAG) and related inference workloads. Built as a subset of the Lenovo Hybrid AI 285 platform, it scales down the 2-8-5 reference configuration — two CPUs, up to eight GPUs, and five network adapters — into a two CPUs, up to eight GPUs, and a single-NIC configuration that is optimized for cost-effective, high-throughput inference at the edge or in the data center.
This guide documents the Lenovo Hybrid AI with 281 with IBM AI software stack, with RAG and Agentic deployment examples from IBM and NVIDIA. The implementation combines Lenovo ThinkSystem SR675 V3 GPU-rich server hardware, NVIDIA RTX Pro 6000 Blackwell Server Edition GPUs, IBM OpenRAG software, IBM Docling document ingestion, IBM Langflow visual flow orchestration, NVIDIA Blueprints, NVIDIA AI Enterprise, OpenSearch vector database, Langfuse, and LiteLLM model routing into a validated, production-ready RAG, Video Intelligence, and Agentic Reserach stack.
The audience for this document includes sales architects, solution engineers, customers, and partners who need a high-level design reference for deploying a cost efficient AI inference platform on Lenovo Hybrid AI infrastructure.
The platform documented here is a single-node AI appliance that can run an entry level, proof of concept, or demo stack such as enterprise document intelligence, video intelligence, and agentic research. All three applications share the same OpenRAG backend, OpenSearch vector store, NVIDIA NIM microservices, LiteLLM model gateway, Langflow orchestration engine, and Langfuse observability layer. The stack can be scaled up for production deployment. Customers can compose, modify, and deploy custom workflows through the same visual Langflow environment without duplicating infrastructure or standing up new integrations.
Table of Contents
1 Introduction
2 Solution Components
3 Solution Architecture
4 Retrieval Augmented Generation (RAG) Functional Test
5 Product Considerations
6 Appendix A: Bill of Materials (BoM)
9 Appendix B: Software Deployment Outline
10 Resources
To view the document, click the Download PDF button.
Configure and Buy
Please select a locale
Full Change History
Course Detail
Employees Only Content
The content in this document with a is only visible to employees who are logged in. Logon using your Lenovo ITcode and password via Lenovo single-signon (SSO).
The author of the document has determined that this content is classified as Lenovo Internal and should not be normally be made available to people who are not employees or contractors. This includes partners, customers, and competitors. The reasons may vary and you should reach out to the authors of the document for clarification, if needed. Be cautious about sharing this content with others as it may contain sensitive information.
Any visitor to the Lenovo Press web site who is not logged on will not be able to see this employee-only content. This content is excluded from search engine indexes and will not appear in any search results.
For all users, including logged-in employees, this employee-only content does not appear in the PDF version of this document.
This functionality is cookie based. The web site will normally remember your login state between browser sessions, however, if you clear cookies at the end of a session or work in an Incognito/Private browser window, then you will need to log in each time.
If you have any questions about this feature of the Lenovo Press web, please email David Watts at dwatts@lenovo.com.
