Cisco and Supermicro Expand Secure AI Factory Collaboration, Shipping October 2026

en.Wedoany.com Reported - On August 25, Cisco expanded its Secure AI Factory with NVIDIA portfolio, partnering with Supermicro to introduce rack-scale high-density AI computing systems. The new solutions cover both liquid-cooled and air-cooled configurations, and include the NVIDIA NVL72 rack-scale system as well as high-density GPU servers based on NVIDIA HGX and MGX architectures. The relevant Supermicro computing systems are planned to be available for ordering through Cisco authorized channels starting October 2026.

The expanded rack-scale architecture adopts configurations that meet the requirements of the NVIDIA Cloud Partner reference architecture. On the compute side, it integrates Supermicro GPU systems; on the network side, it is equipped with Cisco AI Networking, covering front-end, back-end, and storage networks. It also integrates Cisco's distributed security system, the Splunk observability platform, and NVIDIA AI Enterprise software. Storage systems and the Kubernetes platform are provided by Cisco ecosystem partners.

The Supermicro equipment entering the Cisco Secure AI Factory portfolio includes liquid-cooled and air-cooled high-density GPU servers, covering both NVIDIA HGX and MGX platforms. Supermicro's existing NVIDIA AI Factory portfolio also includes HGX B200, HGX B300, and RTX PRO 6000 Blackwell Server Edition systems, with some clusters supporting up to 32 nodes and 256 GPUs per scalable unit.

Cisco is simultaneously advancing Cisco Validated Infrastructure Services, used for configuration validation of AI infrastructure after delivery, and conducting system certification in accordance with NVIDIA Infrastructure Services requirements. The company is also building large-scale AI laboratories for testing related software, tools, and infrastructure. Cisco's existing Secure AI Factory portfolio retains the Cisco UCS computing platform, with UCS products now covering NVIDIA HGX, MGX, and RTX PRO Blackwell series GPU configurations.