An NVIDIA HGX Rubin NVL8 8-GPU system in a 2U direct liquid-cooled chassis, sold and supported by Cisco and managed through Cisco Nexus One. Packing 2.3 TB of HBM4 and NVLink 6 into a quarter of the rack space of the previous generation air-cooled server.

This server brings the NVIDIA Rubin generation to the Cisco AI compute portfolio in its densest form. The underlying hardware is the Supermicro GPU SuperServer SYS-222GS-R2BT-ALC, a 2U liquid-cooled NVIDIA HGX Rubin NVL8 system built on the NVIDIA MGX rack architecture with Supermicro blind-mate busbar and coolant manifold connections for tool-free rack integration. As a rack-integrated liquid-cooled system, it is offered as part of a complete rack with onsite integration service. Cisco handles the selling motion, support through Cisco CX Services, and (Future) lifecycle management through Cisco Intersight. Supermicro builds and integrates the hardware.

The system pairs dual x86 processors with an NVIDIA HGX Rubin 8-GPU connected over sixth-generation NVLink at 3.6 TB/s per GPU, 28.8 TB/s across the baseboard. Eight Rubin GPUs carry 2.3 TB of HBM4 with 176 TB/s of aggregate memory bandwidth and deliver up to 400 petaflops of NVFP4 compute per system. Nine systems fit in one rack for 72 NVLinkconnected Rubin GPUs. Direct liquid cooling is inherent to the platform; for information on any liquid-cooling components, or for a quote on liquid cooling products specifically, contact your Cisco sales specialist.

Table 1 summarizes the primary features and benefits of the platform

Table 1. Features and benefits

Prominent feature

Eight Rubin GPUs in 2U: generational density that air cooling cannot reach.

One 2U NVIDIA HGX NVL8 node carries the same eight-GPU NVLink domain as an 8U air-cooled NVIDIA HGX B300 system, with 2.3 TB of HBM4 instead of HBM3e, three times the memory bandwidth, double the NVLink bandwidth, and roughly four times the NVFP4 training compute. Nine of them stack into a single rack for 72 NVIDIA Rubin GPUs where the previous generation fit far fewer.

That density only works with liquid cooling. Direct liquid cooling carries the compute heat load through blind-mate manifolds, and power arrives over a rack busbar rather than per-node cords, so the system can be sold as part of a complete integrated rack with onsite service rather than as a standalone box. Cooling infrastructure is scoped per facility: in-row and in-rack CDU options and a liquid-to-air sidecar exist for data centers without facility water. For information on any liquid-cooling components, or for a quote on liquidcooling products specifically, contact your Cisco sales specialist.

Cisco offers this Supermicro platform in a defined configuration, given in Table 2. Supermicro has not yet published the full public data sheet for this SKU; the fields in Table 2 reflect the announced 2U HGX Rubin NVL8 platform.

Table 2. Configuration selected by Cisco, SYS-222GS-R2BT-ALC (HGX Rubin NVL8 liquid-cooled)

Information about Cisco’s Environmental, Social and Governance (ESG) initiatives and performance is provided in Cisco’s CSR and sustainability reporting.

Table 3. Cisco Environmental Sustainability Information

Table 4 reflects the NVIDIA HGX R200 NVL8 platform and Supermicro announcements for the 2U liquid-cooled system. Supermicro has not yet published the full data sheet for this SKU; product specifications may change without notice.

Table 4. Product specifications, Supermicro GPU SuperServer SYS-222GS-R2BT-ALC

Table 5 lists what must be in place before installation. This is a rack-integrated liquid-cooled system; for information on any liquid-cooling components, or for a quote on liquid-cooling products, contact your Cisco sales specialist.

Table 5. System requirements

The base platform is a Supermicro SKU offered through Cisco as part of an integrated rack. Liquid-cooling components (coolant distribution units, manifolds, and liquid-to-air sidecars) are quoted separately: contact your Cisco sales specialist for information on any liquid-cooling components or for quotes on specific liquid-cooling products. To order, contact your Cisco account representative or visit the Cisco Ordering Home Page.

Table 6. Ordering information

Warranty information

Warranty terms for this Supermicro platform offered by Cisco are standard Supermicro warranty terms. Support is delivered through Cisco CX services under the applicable service contract; contact your Cisco sales specialist for more details.

Cisco and Partner Services

Cisco CX services covers this platform from planning through deployment and ongoing operations, with a single Cisco support contract even though the hardware is Supermicro-built. For more information, visit https://www.cisco.com/go/services.

Cisco Capital

Flexible payment solutions to help you achieve your objectives

Cisco Capital makes it easier to get the right technology to achieve your objectives, enable business transformation and help you stay competitive. We can help you reduce the total cost of ownership, conserve capital, and accelerate growth. In more than 100 countries, our flexible payment solutions can help you acquire hardware, software, services and complementary third-party equipment in easy, predictable payments. Learn more. 

Learn more

For more on Cisco AI infrastructure, visit https://www.cisco.com/go/ai. Full platform specifications will be available on the Supermicro product page for the SYS-222GS-R2BT-ALC once published.

Learn more about Cisco and Supermicro partnership:
https://www.cisco.com/site/us/en/solutions/global-partners/supermicro/index.html