FusionServer G6500 V8
AI Server

Leading Computing, Flexible Configuration,
Simplified Architecture

  • Home
  • FusionServer V8
  • FusionServer G6500 V8
    5288 V7超大存储

    Leading Computing

    5288 V7可靠安全

    Flexible Configuration

    5288 V7高效节能

    Simplified Architecture

    Next-Gen Flagship Air-Cooled AI Server

    Major application scenarios:

    AI training

    AI inference

    HPC

    The FusionServer G6500 V8 is a next-generation flagship air-cooled AI server engineered to meet the rigorous demands of large model training and inference.

    Featuring exceptional performance, robust reliability, and ease of deployment and maintenance, the G6500 V8 accelerates AI training and inference, High-Performance Computing (HPC), image/video analysis, and database applications. With flexible configurations and scalable AI computing, the G6500 V8 stands as a powerful platform for driving AI innovation and transforming industries.

    Next-Gen Flagship Air-Cooled AI Server

    Major application scenarios:

    AI training

    AI inference

    HPC

    The FusionServer G6500 V8 is a next-generation flagship air-cooled AI server engineered to meet the rigorous demands of large model training and inference.
    Featuring exceptional performance, robust reliability, and ease of deployment and maintenance, the G6500 V8 accelerates AI training and inference, High-Performance Computing (HPC), image/video analysis, and database applications. With flexible configurations and scalable AI computing, the G6500 V8 stands as a powerful platform for driving AI innovation and transforming industries.
    5288 V7新一代4U2路机架服务器

    Leading Computing

    Exceptional AI Computing

    20 x single-width GPUs

    10 x dual-width GPUs

    Powerful CPU Performance

    2 x Intel®Xeon®6 processors, each with up to 350 W TDP

    Exceptional AI Computing

    20 x single-width GPUs

    10 x dual-width GPUs

    Flexible Configuration

    Flexible Configuration

    Multiple GPU form factors, with a maximum TDP of 600 W
    Flexible options and configurations of pass-through and switch topologies

    Simplified Architecture

    Modular Design

    Seven major modules* with high convergence and low coupling, allowing for
    flexible reconfiguration and independent evolution
    *Seven modules refer to GPU, CPU, I/O, power, fan, drive, and chassis modules.

    Flexible Cabling

    Front or rear cable routing options based on actual data center scenarios and maintenance practices

    Efficient O&M

    Cable blind-mating design; maintenance without component removal

    Contact Us

    Contact Sales

    Contact Sales

    Please fill in the form below, we will contact you as soon as
    possible.