# Lenovo ThinkSystem SR670 V2/V3 Series
> The Lenovo ThinkSystem SR670 V2/V3 Series is a family of high-performance rack servers optimized for artificial intelligence, high-performance computing (HPC), and accelerated graphics workloads. These systems feature modular designs to support high-density GPU configurations, including NVIDIA HGX and PCIe adapters, often incorporating advanced cooling technologies like Lenovo Neptune.
- URL: https://optimly.ai/brand/lenovo-thinksystem-sr670-v2-v3
- Logo: https://logo.clearbit.com/lenovo.com
- Slug: lenovo-thinksystem-sr670-v2-v3
- BAI Score: 88/100
- Archetype: Challenger
- Category: Hardware
- Last Analyzed: July 19, 2026
## Buyer Intent Signals
Problems: Status Quo / Legacy Infrastructure: Continuing to run legacy, non-GPU optimized rack servers for AI workloads, leading to high latency and energy inefficiency. | Custom Infrastructure Engineering: Hiring specialized data center consultants to custom-build liquid-cooled HPC clusters from white-box components.
Solutions: best server for training large language models | NVIDIA HGX H100 server manufacturers | liquid cooled GPU server for enterprise | highest density GPU rack server | Public Cloud Infrastructure (IaaS): Utilizing cloud-based GPU instances (e.g., AWS P4d, Azure NDv4) instead of owning physical hardware.
Comparisons: ThinkSystem SR670 V3 vs V2 specs
---
## Full Details / RAG Data
### Overview
Lenovo ThinkSystem SR670 V2/V3 Series is listed in the AI Directory.
The Lenovo ThinkSystem SR670 V2/V3 Series is a family of high-performance rack servers optimized for artificial intelligence, high-performance computing (HPC), and accelerated graphics workloads. These systems feature modular designs to support high-density GPU configurations, including NVIDIA HGX and PCIe adapters, often incorporating advanced cooling technologies like Lenovo Neptune.
### Metadata
| Field        | Value |
|--------------|-------|
| Name         | Lenovo ThinkSystem SR670 V2/V3 Series |
| Slug         | lenovo-thinksystem-sr670-v2-v3 |
| URL          | https://optimly.ai/brand/lenovo-thinksystem-sr670-v2-v3 |
| Logo         | https://logo.clearbit.com/lenovo.com |
| BAI Score    | 88/100 |
| Archetype    | Challenger |
| Category     | Hardware |
| Last Analyzed | July 19, 2026 |
| Last Updated | 2026-07-25T06:57:34.729Z |
### Verified Facts
- Founded: 2017
- Headquarters: Morrisville, North Carolina, USA and Beijing, China
### Buyer Intent Signals
#### Problems this brand solves
- Status Quo / Legacy Infrastructure: Continuing to run legacy, non-GPU optimized rack servers for AI workloads, leading to high latency and energy inefficiency.
- Custom Infrastructure Engineering: Hiring specialized data center consultants to custom-build liquid-cooled HPC clusters from white-box components.
#### Buyers search for
- best server for training large language models
- NVIDIA HGX H100 server manufacturers
- liquid cooled GPU server for enterprise
- highest density GPU rack server
- Public Cloud Infrastructure (IaaS): Utilizing cloud-based GPU instances (e.g., AWS P4d, Azure NDv4) instead of owning physical hardware.
#### Buyers compare
- ThinkSystem SR670 V3 vs V2 specs
### Links
- Canonical page: https://optimly.ai/brand/lenovo-thinksystem-sr670-v2-v3
- JSON endpoint: /brand/lenovo-thinksystem-sr670-v2-v3.json
- LLMs.txt: /brand/lenovo-thinksystem-sr670-v2-v3/llms.txt