AnswersPublished21 sources
Cisco Adds Supermicro AI Racks to Secure AI Factory in Push Beyond Networking
On August 25, 2026, Cisco announced that Supermicro’s liquid and air cooled rack scale systems will join its Secure AI Factory with NVIDIA, with pre validated solutions planned for October. The systems are designed for enterprise, neocloud, and sovereign cloud deployments and will support NVIDIA Vera Rubin NVL72 and...
Published byEdited with GPT-5.6 Luna
Research answer
Key takeaways
- On August 25, 2026, Cisco announced that Supermicro’s liquid and air cooled rack scale systems will join its Secure AI Factory with NVIDIA, with pre validated solutions planned for October.
- The systems are designed for enterprise, neocloud, and sovereign cloud deployments and will support NVIDIA Vera Rubin NVL72 and HGX Rubin NVL8 platforms for workloads ranging from very large model training to high thr...
- Cisco will bundle the compute with its networking, security, observability, and deployment services, while adding Cisco Validated Infrastructure Services and plans for a dedicated large scale AI lab.
AI Prompt
microsoft.comCreate a landscape editorial hero image for this Studio Global article: What did Cisco announce on August 25, 2026, about its partnership with Supermicro to expand Cisco’s Secure AI Factory with NVIDIA portfolio,. Article summary: Cisco announced that it will add Supermicro’s high-density, liquid- and air-cooled rack-scale compute systems to its Secure AI Factory with NVIDIA, turning Cisco’s offering into a more complete, pre-validated AI-infrastr. Topic tags: general, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fa
Cisco is expanding its Secure AI Factory with NVIDIA beyond networking and security. Through a new partnership with Supermicro announced on August 25, 2026, Cisco plans to add high-density rack-scale systems with both liquid and air cooling to its AI-infrastructure portfolio. The pre-validated systems are expected to become available through Cisco’s channel in October 2026. 1
2
8
The announcement is strategically important because it gives Cisco a larger role in assembling and validating the infrastructure required to run modern AI clusters. Supermicro supplies the dense compute and cooling systems; NVIDIA supplies the GPU platforms; and Cisco combines them with networking, security, observability, and deployment services.
What Cisco and Supermicro announced
Cisco said it will offer Supermicro rack-scale and dense-GPU systems as part of the broader Secure AI Factory with NVIDIA architecture. The portfolio is aimed at enterprises, neocloud providers, and sovereign-cloud operators, in addition to larger cloud and hyperscale environments. 7
8
9
The planned configurations include support for:
- NVIDIA Vera Rubin NVL72, a rack-scale platform designed for large AI deployments.
- NVIDIA HGX Rubin NVL8, a dense GPU platform for AI infrastructure deployments.
- Liquid- and air-cooled Supermicro systems, allowing customers to select different approaches to managing the thermal demands of high-density compute.
Cisco is positioning the expanded architecture for workloads that include very large-model or trillion-parameter training and high-throughput inference. Those workload descriptions come from the companies and their reporting partners, so they describe the intended use cases rather than independently verified performance results. 2
9
14
What customers are supposed to buy
The partnership is less about selling a standalone server than about packaging multiple infrastructure layers into a validated system. Customers will be able to procure Supermicro compute alongside Cisco AI networking, security, observability, and deployment services through Cisco’s authorized channel. 1
5
That approach addresses a practical problem in AI data-center construction: organizations must coordinate GPUs, servers, cooling, networking, security controls, software, and operations. A pre-validated architecture can reduce the amount of testing and integration work required before a cluster is put into production, although the announcement does not quantify the time or cost savings.
Cisco’s existing architecture includes Cisco networking and security products, with AI networking that spans Cisco Silicon One and NVIDIA Spectrum-X Ethernet silicon. 6
11 The expanded offering therefore gives Cisco a way to remain central to the network and operational layers while adding more of the compute infrastructure around them.
Cisco’s networking differentiation claim
Cisco said its NVIDIA Cloud Partner-compliant architecture is distinct because it uses Cisco’s own networking switches and network operating system. 9
That is a Cisco product-positioning claim, not an independently verified conclusion that no other NVIDIA technology partner uses its own switching hardware or network operating system. The more supportable takeaway is that Cisco is emphasizing control over the networking layer as a differentiator in an otherwise NVIDIA-centered AI stack.
Validation and operations become part of the offer
Cisco also announced Cisco Validated Infrastructure Services, intended to help customers certify that their systems are designed and aligned with the relevant reference architecture. The company said the services will be aligned with NVIDIA Infrastructure Services. 7
14
Cisco further said it plans to establish a dedicated large-scale AI lab. The purpose is to validate rack-scale designs and help customers deploy and operate them more reliably, according to the company’s announcement and related coverage. 8
11
This operational layer matters because rack-scale AI systems are not simply collections of interchangeable servers. Power, cooling, fabric design, software configuration, security, and ongoing observability all affect whether a large cluster works as intended. Cisco is using validation and services to extend its role beyond selling switches and into the lifecycle of the AI infrastructure.
Why the deal matters to Cisco’s strategy
Cisco’s core strategic move is to become a broader infrastructure control point for AI data centers. Rather than competing directly with NVIDIA as a GPU designer or replacing hyperscalers as cloud operators, Cisco is seeking to own more of the integration layer around those platforms.
The Supermicro partnership fills a gap in Cisco’s portfolio: high-density compute and cooling for rack-scale systems. Cisco contributes the networking, security, observability, channel reach, and validation capabilities that can make the combined system easier for customers to procure and operate. 5
8
The target market is also broader than hyperscalers. Neoclouds and sovereign-cloud providers may want to deploy large AI environments while retaining more control over infrastructure, location, and operations. Cisco explicitly named those groups alongside enterprises in describing the expanded portfolio. 7
8
The growth backdrop—and what the announcement does not disclose
The expansion follows a sharp increase in Cisco’s reported AI-infrastructure demand. Cisco’s fiscal 2026 materials cited approximately $9.3 billion in hyperscaler AI-infrastructure orders, about 4.5 times the prior year, and expectations for approximately $7.5 billion in related fiscal 2027 revenue. 20
However, the Supermicro announcement itself did not disclose a customer purchase, rack count, contract value, revenue-sharing arrangement, or margin for either company. 1 That limits what can be concluded about the partnership’s near-term financial impact.
Market coverage also cited different stock-performance figures. The specific claim of an approximately 45% year-to-date gain was not substantiated by the sources reviewed; one report instead measured a 68.5% gain for Cisco over the year ending August 25, 2026, compared with 19.7% for the S&P 500 over the same period. 21 Stock performance should therefore be treated as market context, not evidence that the partnership has already produced customer revenue.
The bottom line
Cisco’s August 25 announcement adds Supermicro’s liquid- and air-cooled rack-scale compute to the Secure AI Factory with NVIDIA and broadens the offer from networking-centered infrastructure to a more complete, pre-validated AI system.
The planned October 2026 rollout is aimed at customers building large training and inference environments on NVIDIA Vera Rubin platforms. Cisco’s opportunity is to simplify the path from GPU procurement to an operational AI cluster by combining compute, cooling, networking, security, observability, and validation.
The key test will come after launch: whether customers view Cisco’s integrated architecture as valuable enough to justify buying through one broader infrastructure stack instead of assembling servers, GPUs, cooling, and network systems independently.
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "Cisco Adds Supermicro AI Racks to Secure AI Factory in Push Beyond Networking"?
On August 25, 2026, Cisco announced that Supermicro’s liquid and air cooled rack scale systems will join its Secure AI Factory with NVIDIA, with pre validated solutions planned for October.
What are the key points to validate first?
On August 25, 2026, Cisco announced that Supermicro’s liquid and air cooled rack scale systems will join its Secure AI Factory with NVIDIA, with pre validated solutions planned for October. The systems are designed for enterprise, neocloud, and sovereign cloud deployments and will support NVIDIA Vera Rubin NVL72 and HGX Rubin NVL8 platforms for workloads ranging from very large model training to high thr...
What should I do next in practice?
Cisco will bundle the compute with its networking, security, observability, and deployment services, while adding Cisco Validated Infrastructure Services and plans for a dedicated large scale AI lab.
Sources
r40.ioCisco Added Supermicro to Its AI Factory. | R40
superpowerdaily.comCisco Adds Supermicro Rack-Scale Systems to Its NVIDIA ...
newsroom.cisco.comCisco Expands Secure AI Factory with NVIDIA for Rack-Scale Era
siliconangle.comCisco expands rack-scale secure AI factory infrastructure for neocloud and sovereign clouds - SiliconANGLE