Cisco UCS C480 ML M5 Rack Server - Custom Build to Order

The Cisco UCS C480 ML M5, also known by the model number UCSC-C480-ML-M5, is a custom build-to-order server, high-performance, 4-socket, 4U rack server specifically designed for deep learning workloads. It offers exceptional processing power, storage flexibility, and I/O optimization to handle the demanding requirements of training complex AI models.

Here are some key features and options you can choose from when configuring your Cisco UCS C480 ML M5 server:

• Processors: Supports up to four 2nd Generation Intel® Xeon® Scalable processors, delivering the necessary processing power for intricate calculations involved in deep learning.
• Memory: Up to 24 DIMM slots, allowing for a maximum memory capacity of 7.5 TB using DDR4 DIMMs.
• Storage:
  • Up to 24 front-loading 2.5-inch SAS/SATA drives for bulk storage.
  • Eight NVIDIA SXM2 V100 32GB modules with NVLink interconnect: This is a key feature for deep learning, providing high-performance processing and communication capabilities specifically suited for AI workloads.
  • Up to six PCIe NVMe disk drives for additional high-speed storage needs.
• Networking:Power: Redundant 2800W or 3000W power supplies to ensure system uptime and handle the power demands of AI workloads.
  • Two embedded 10 Gigabit Ethernet (10GbE) LAN ports for basic connectivity.
  • Modular LAN on motherboard (mLOM) slot for additional high-speed networking options.
• Power: Redundant 2800W or 3000W power supplies to ensure system uptime and handle the power demands of AI workloads.

Advantages of Cisco UCS C480 ML M5:

• Purpose-built for deep learning: The combination of powerful processors, NVIDIA SXM2 GPUs, and optimized storage caters specifically to the needs of training complex AI models.
• Performance and scalability: The customizable nature allows you to tailor the server to your specific deep learning tasks, scaling up as your requirements evolve by adding more memory, storage, or networking components.
• I/O flexibility: With multiple PCIe slots and support for various storage configurations, the server caters to diverse I/O requirements of deep learning workloads.

Applications:

• Deep learning training and inference
• Artificial intelligence (AI) development and research
• Scientific computing
• High-performance computing (HPC)