SuperServer SRS-48UAC-RTXPRO

AI Factory Solutions with NVIDIA RTX PRO 6000 Blackwell Server Edition

Key Applications

  • Artificial Intelligence (AI)
  • Design & Visualization
  • Deep Learning Inferencing
  • GPU Virtualization (VDI)
  • LLMs & Multi-modal LLMs

Key Features

  1. Full-stack solutions based on reference architectures that include Supermicro systems, NVIDIA GPUs, NVIDIA software, and NVIDIA networking
  2. Up to 256x NVIDIA RTX PRO™ 6000 Blackwell Server Edition GPUs providing up to 24TB GDDR7 memory
  3. Compatibility with the NVIDIA software stack (NVIDIA AI Enterprise, NVIDIA Omniverse™, and NVIDIA Run:ai)
  4. Plug-and-play solution with systems fully integrated into racks and tested before shipment and on-site deployment
  5. NVIDIA Spectrum™-X Ethernet Compute Fabric, Converged Network, and Out-of-Band Management included
  6. Supermicro AI Factory solutions are endorsed by NVIDIA for Infrastructure Configuration, Spectrum-X networking, and Software Reference Stack and based on the NVIDIA Enterprise Reference Architecture for RTX PRO 6000 Blackwell Server Edition.
...
...
Cluster Overview
Compute Nodes
  • 4, 8, or 32 nodes (32–256 GPUs) per scalable unit
Compute Racks
  • Up to 4x compute racks per scalable unit
Networking Racks
  • Up to 1x networking rack per scalable unit
Rack Form Factor
  • 48U 19"-width rack enclosure
Rack Dimensions
  • 750mm x 1200mm
Total GPUs
  • Up to 256x NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs
Total GPU Memory
  • Up to 24TB GDDR7
Compute Node Configuration
System SYS-422GL-NR
GPU
  • 8x NVIDIA RTX PRO 6000 Blackwell Server Edition GPU
CPU
  • 2x Intel® Xeon® 6972P Processor 96-Core 2.40GHz 480MB Cache (400W)
Memory
  • 24x 64GB DDR5 6400MHz ECC RDIMM
Local Node Storage
  • Micron 7450 PRO 960GB NVMe
Networking
  • 1x NVIDIA BlueField®-3 (B3220) and 4x NVIDIA BlueField-3 (B3140H)
Power
  • 9.1kW (full load power draw)
System SYS-522GA-NRT
GPU
  • 8x NVIDIA RTX PRO 6000 Blackwell Server Edition GPU
CPU
  • 2x Intel® Xeon® 6972P Processor 96-Core 2.40GHz 480MB Cache (400W)
Memory
  • 24x 64GB DDR5 6400MHz ECC RDIMM
Local Node Storage
  • Micron 7450 PRO 960GB NVMe
Networking
  • 1x NVIDIA BlueField®-3 (B3220) and 4x NVIDIA BlueField-3 (B3140H)
Power
  • 9.1kW (full load power draw)
System AS -5126GS-TNRT2
GPU
  • 8x NVIDIA RTX PRO 6000 Blackwell Server Edition GPU
CPU
  • 2x AMD EPYC™ 9575F Processor 64-Core 3.3GHz 256MB Cache (400W)
Memory
  • 24x 64GB DDR5 6400MHz ECC RDIMM
Local Node Storage
  • Samsung PM9A3 960GB NVMe PCIe
Networking
  • 1x NVIDIA BlueField®-3 (B3220) and 4x NVIDIA BlueField-3 (B3140H)
Power
  • 8.6kW (full load power draw)
Compute Rack Configuration
Compute Nodes
  • 4 or 8 Compute Nodes per rack
Estimated Rack Power
  • 4 nodes per rack: ~36.6kW (SYS-522GA-TNR), ~33.3W(SYS-422GL-NR), and ~34.2kW (AS -5126GS-TNRT2). 8 nodes per rack: ~73.2kW (SYS-522GA-TNR), ~66.5W(SYS-422GL-NR), and 68.5kW (AS -5126GS-TNRT2).
Total GPUs
  • 32 GPUs (4 nodes) or 64 GPUs (8 nodes) per rack
Total CPUs
  • 8 CPUs (4 nodes) or 16 CPUs (8 nodes) per rack
Networking Overview
Compute Fabric
  • Up to 2x NVIDIA Spectrum-4 800GbE Ethernet Switch (SN5610)
Converged Network
  • Up to 2x NVIDIA Spectrum-4 800GbE Ethernet Switch (SN5610)
Out-of-Band Management
  • Up to 4x NVIDIA Spectrum 1GbE Ethernet Switch (SN2201)
Software Overview
Infrastructure Software
  • NVIDIA Run:ai
Application Software
  • NVIDIA AI Enterprise
  • NVIDIA Omniverse™
Services
Onsite Deployment
  • Plug-and-play on-site deployment and validation
Rack Integration
  • Full rack-level design, component installation, and cabling
  • System-level testing of each node (L11)
  • Cluster-level multi-node testing (L12)

SuperCluster SRS-48UAC-RTXPRO New

Bestimmte Produkte sind in Ihrer Region möglicherweise nicht verfügbar.