Skip to product information
1 of 1

Cisco

Cisco UCSC-885A-M8-H21 | UCS C885A M8 8x H200 GPU AI Server, 2.3TB, 8x 400G

SKU:UCSC-885A-M8-H21

Request Quote
Sale Sold out
Shipping calculated at checkout.

Stock Status: Enquire

 Trade supply only. Server & GPU quote requests need a company name, ABN and work email address. We do not quote Gmail, Outlook or Hotmail addresses, or supply for home use. 

20-min response

Need a quote or bulk pricing? Answered within 20 minutes, Sydney business hours.

Custom Built Servers

Need a different spec, quantity or configuration?

We build configured-to-order refurbished servers from Cisco, Dell, HPE, NVIDIA Lenovo and Supermicro. Tell us what you need and we'll come back with options within 24 hours.

Configure your server now →

Description

The UCS C885A M8 is an 8RU AI rack server in the Cisco UCS C-Series, built on the NVIDIA HGX platform for LLM training, fine-tuning, large-model inference, retrieval-augmented generation and wider deep learning and HPC cluster workloads. This SKU is a fixed configuration with 8 NVIDIA H200 SXM GPUs connected over NVLink.

Compute and memory are provided by two AMD EPYC 9554 processors (64 cores each, 3.1 GHz, boosting to 3.75 GHz) and 24 x 96GB DDR5 4,800 MT/s RDIMMs, giving 2.3TB of system memory. Local storage is a 1TB M.2 NVMe boot drive plus 16 x 2.5-inch NVMe SSDs for fast data caching. Networking is split by traffic type: 8 x NVIDIA ConnectX-7 1x400G NICs handle East-West traffic, one NVIDIA BlueField-3 B3220 DPU (2x200G) handles North-South traffic, and an OCP 3.0 slot carries an Intel X710-T2L 2x10G RJ45 NIC.

The 400G ConnectX-7 NICs are intended for clustering over 400G fabrics, and Cisco positions the server within its AI POD alongside Nexus 9364E-SG2 switches. Management is through Cisco Intersight.

Enquire with XS Network Tech for new or used pricing and availability.

Features

- 8 NVIDIA H200 SXM GPUs with NVLink interconnect for dense AI training and inference
- Two high-clock-speed AMD EPYC CPUs selected for training and learning workloads
- 8 x 400G East-West ConnectX-7 NICs for chassis-to-chassis and GPU-to-GPU connectivity at scale
- NVIDIA BlueField-3 DPU to speed up GPU access to data
- 16 NVMe SSD bays for fast local data caching
- Hot-swappable redundant power supplies for high availability
- Managed through Cisco Intersight

Warranty

All products sold by XS Network Tech include a 12-month warranty on both new and used items. Our in-house technical team thoroughly tests used hardware prior to sale to ensure enterprise-grade reliability.

All technical data should be verified on the manufacturer data sheets.

View full details

specs-tabs

Collapsible content

Technical Specifications

FAQs

Technical Specifications

- Form factor: 8RU rack server
- Processors: 2 x AMD EPYC 9554, 64 cores each, 3.1 GHz (up to 3.75 GHz boost)
- GPUs: 8 x NVIDIA HGX H200 SXM with NVLink
- Memory: 24 x 96GB DDR5 4,800 MT/s RDIMM (2.3TB total)
- Boot drive: 1 x 1TB M.2 NVMe
- Internal storage: 16 x 2.5-inch U.2 NVMe SSD (platform maximum is 16)
- East-West networking: 8 x NVIDIA ConnectX-7 (1x400G), PCIe Gen5 x16 HHHL
- North-South networking: 1 x NVIDIA BlueField-3 B3220 DPU (2x200G), PCIe Gen5 x16 FHHL
- OCP slot: 1 x OCP 3.0 PCIe Gen5 x8 with Intel X710-T2L 2x10G RJ45 NIC
- PCIe slots (platform): 5 x PCIe Gen5 x16 FHHL for North-South NICs and 8 x PCIe Gen5 x16 HHHL for East-West NICs
- Power supply options (platform): 2 x 2.7kW 80Plus 12V CRPS hot-swap redundant (N+1) or 6 x 3kW 80Plus 54V MCRPS hot-swap redundant (N+2)
- Management: Cisco Intersight
- Encryption: NICs and DPUs ship with encryption disabled

FAQs

Q: Does this server need a licence to operate?
A: No licence is listed for the hardware to operate. Management is through Cisco Intersight, with the local server management interface handling initial configuration and Intersight providing inventory through the embedded device connector.

Q: What power supplies does the platform use?
A: The platform supports either 2 x 2.7kW 80Plus 12V CRPS hot-swap redundant supplies (N+1) or 6 x 3kW 80Plus 54V MCRPS hot-swap redundant supplies (N+2). Confirm the power supply option for your build before ordering.

Q: What networking does this configuration provide?
A: It has 8 x NVIDIA ConnectX-7 1x400G NICs for East-West GPU traffic, one NVIDIA BlueField-3 B3220 DPU (2x200G) for North-South traffic, and an OCP 3.0 Intel X710-T2L 2x10G RJ45 NIC.

Q: What does it work with?
A: The ConnectX-7 400G NICs are intended for clustering with 400G fabrics, and Cisco positions the server in its AI POD with Nexus 9364E-SG2 switches. Supported operating systems and peripherals are listed in the Cisco Hardware and Software Interoperability List.

Q: How does this SKU differ from other C885A M8 configurations?
A: This SKU is the H200 GPU build with 2 x AMD EPYC 9554, 2.3TB of DDR5 memory, 8 x ConnectX-7 1x400G NICs, one BlueField-3 B3220 DPU, a 1TB boot drive and 16 NVMe SSDs.

  • Request a Quote

    Looking for competitive pricing? Submit a request, and our team will provide a tailored quote that fits your needs.

  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Call us Now  
  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Contact us