2025 Computex Taipei
2025 Computex Taipei
  • Date
    May 20 - 23, 2025
  • Location
    Taipei Nangang Exhibtion Center, Hall 1, Floor 4
  • Booth
    L0106
Share

GB12481N

HGX B200 8-GPU Al Solution Optimized for Accelerated Computing and Generative AI
Support NVIDIA HGX B200 8-GPU
Deploy NVIDIA NVLink® & NVSwitch™ Technology
Two-layer GPU and I/O Sleds for Easy Scalability and Better Thermal Efficiency
Support 20 x PCIe 5.0 Slots & 32 x NVMe Drive Bays
Air Cooling Solution
Powered by
NVIDIA

Application

Generative AI

Generative AI

Large Language Model (LLM)

Large Language Model (LLM)

Hyperscale Data Center

Hyperscale Data Center

High Performance Computing

High Performance Computing

AI and Machine Learning

AI and Machine Learning

hga_h100

HGX B200 8-GPU Al Solution

Built on the NVIDIA Blackwell architecture, this AI server is optimized for generative AI and high-performance computing. It integrates eight NVIDIA B200 Tensor Core GPUs, connected via fifth-generation NVIDIA NVLink, delivering up to 72 petaFLOPS for training and 144 petaFLOPS for inference. With a total of 1.4 TB of HBM3e memory across the GPUs, the system provides exceptional memory bandwidth, enabling efficient processing of large-scale generative AI, data analytics, and HPC workloads.
 
 
powerful (1)

Innovative Disaggregated Approach

The AI server can be paired with a single CPU head node server, creating a unified AI solution. This disaggregated GPU and CPU approach enhances flexibility, enabling you to select the CPU head node server that best suits your needs.
(Ingrasys head node solution: SV2221A/SV2221I)
 

 

hga_exp

High Expansion Capability

Adopting a one-GPU-to-one-NIC topology, the powerful server offers high expansion capability by accommodating 8 NICs, which enables fast interconnections between GPUs within a GPU cluster. In addition, its two-layer design features easy scalability up to 32 NVMe Drives and enhances thermal efficiency, allowing the server to deliver unprecedented performance and agility.

Certification

FCC
RCM
BSMI
CB
CE
UL
Supported GPU

NVIDIA HGX B200 8-GPU

Expansion Slots 20 x PCIe 5.0 x16 Slots
Storage 32 x Hot-swap U.2 NVMe Drive Bays
Front Panel

1 x Power LED, 1 x UID LED, 1 x Attention LED
1 x 1GbE/RJ45 Management Port
1 x RJ45 Console Port

Form Factor 12 U Rackmount
(4U IOB Sled with 6U GPU Sled)
Chassis Dimensions
(H x W x D)
20.8" x 20.0"x 37.3"/ 529.4mm x 508.0mm x 948.0mm

Management

1 x ASPEED AST2600
Power Supply 6+6 Redundant 3000W Platinum Power Supplies 
Fans 30 x 80*80mm for N+1 Cooling Redundancy
Certification CE/FCC/RCM/BSMI/UL/IECEE CB

Operating Temperature

10°C to 35°C (50°F to 95°F)
Non-operating Temperature     -40°C to 60°C (-40°F to 140°F)

Operating Relative Humidity 

8% to 85%RH
Non-operating Relative Humidity 5% to 95%RH
TOP

We use cookies to allow our website to work properly, personalize content and advertising, provide social media features and analyze traffic. We also share information about your use of our site with our social media, advertising and analytics partners

Manage Cookies

Privacy preferences

We use cookies to allow our website to work properly, personalize content and advertising, provide social media features and analyze traffic. We also share information about your use of our site with our social media, advertising and analytics partners

Privacy Policy

Manage preferences

Necessary cookie

Always on

The website cannot function without these cookies and you cannot switch them off on your system. These cookies are typically set solely in response to an action you perform (i.e. a service request), such as setting privacy preferences, logging in, or filling in a form. You can set your browser to block or prompt you for these cookies, but this may prevent some site features from working.