Pergunte aos nossos especialistas

Soluções de Projeto

Obter aconselhamento: Bate-papo ao vivo | +1-626-655-0998 | E-mail

AI Server

AI servers are designed for GPU-accelerated workloads such as model training, inference, fine-tuning, deep learning, computer vision, and high-performance computing. When selecting an AI server, users typically focus on GPU quantity, GPU memory, CPU and system memory, GPU interconnect, storage bandwidth, networking, power, and cooling.

The comparison below groups representative AI server models by GPU configuration. Use it to identify a suitable server scale before reviewing the complete AI server product list, exact configurations, availability, and pricing.

AI Server Models by GPU Configuration

Compare representative AI servers by GPU capacity, system configuration, and typical workload.

GPU Configuration Representative AI Server Models Configuration Highlights Recommended For
1–2 GPU AI Servers AT600W
AT600WDD
1–2 Atlas 300I Duo accelerators, compact system design, and moderate CPU and memory capacity for localized AI processing. AI inference, model development, edge AI, and smaller computer vision workloads.
4 GPU AI Servers AT800
AT9508 G3
Four Atlas 300I Duo accelerators, higher memory and PCIe capacity, and greater throughput for parallel AI workloads. Model fine-tuning, computer vision, analytics, and medium-scale inference.
8 GPU AI Servers xFusion G530 V5
H200 AI Server Bundle11
H100 SYS-821GE-TNHR
Nvidia H200 AI Bundle10
Eight high-end GPUs, large GPU memory capacity, high-speed GPU interconnects, expanded system memory, and high-bandwidth networking. Large-model training, deep learning, distributed inference, generative AI, and HPC workloads.
High-Density GPU Systems DTG-6702-X5
DTG-6802-X5
DTG-7602-X5
Support for multiple high-power or wide GPUs, large PCIe expansion capacity, high-capacity power supplies, and enhanced airflow. AI clusters, large-scale training, GPU cloud infrastructure, scientific computing, and high-performance accelerated applications.

Before ordering, verify the exact GPU model and memory, CPU and system memory, power supplies, cooling requirements, rack space, network connectivity, and software compatibility of the selected configuration.

Explore More AI Computing Resources

Explore GPU Servers

Explore HPC Servers

Filtrar Por

Itens 1-20 de 51

Definir Ordenação Decrescente
Página
Ver como Lista Grelha
  1. H100 SYS-821GE-TNHR, Supermicro GPU Server, 2x8468 CPUs/32x64GB DDR5/8xH100 GPUs

    Condição: Original Novo

    2*8468+32*64G\2*960GB NVMe SSD+8*3.84TGB NVMe SSD\8*H100
    302 999,00 US$
  2. H100 SYS-821GE-TNHR, Supermicro H100 GPU Server, 2x8468/32x64GB RAM/4x3.84TB NVMe/8xH100

    Condição: Original Novo

    2*8468+32*64G\4*3.84TGB NVMe SSD\8*H100
    299 999,00 US$
  3. HGX SYS-821GE-TNHR, Supermicro HGX Server, 2xIntel 8558+, 8x64GB DDR5, 8xNVIDIA Delta-Next HGX GPU

    Condição: Original Novo

    2*8558+\8*64G\960GB SSD\8*NVIDIA DELTA-NEXT HGX GPU
    297 999,00 US$
  4. H100 SYS-821GE-TNHR, Supermicro H100 GPU Server, 2x8480+ CPUs/32x64GB RAM/8xH100 GPU

    Condição: Original Novo

    2*8480+\32*64G\2*960GB NVME SSD+2*15.6TB NVME SSD\8*H100
    309 999,00 US$
  5. H200 SYS-821GE-TNHR, Supermicro H200 Series GPU Server, 2x8468/32x64G DDR5/8xH200 GPU/2x7.6TB NVMe SSD

    Condição: Original Novo

    2*8468/32*64G DDR5/2*M.2 NVMe 960G+2*7.6TB NVMe SSD\8*H200
    314 299,00 US$
  6. H200 SYS-821GE-TNHR, Supermicro H200 GPU Server, 2x8558/32x64G/8xH200

    Condição: Original Novo

    2*8558/32*64G/M.2 NVMe 960G+4*3.84TB NVMe/8*H200
    314 299,00 US$
  7. H100 SYS-821GE-TNHR, Supermicro H100 GPU Server, 2x8558/32x64G/8xH100

    Condição: Original Novo

    2*8558/32*64G/2*960G SATA +8*3.84T/8*H100
    298 999,00 US$
  8. SYS-821GE-TNHR, Supermicro SYS-821GE Series, GPU Server, 2x8468/32x64G DDR5/4x3.8TB NVMe/8xH100

    Condição: Original Novo

    2*8468/32*64G DDR5/480G SATA+4*3.8TB NVMe\8*H100
    298 999,00 US$
  9. Nvidia H100 AI Bundle 9, Inspur Nvidia H100 Series, 2x8480 CPU/32x64GB RAM/8xH100 GPU

    Condição: Original Novo

    2*8480+\32*64G\2*1.92TB M.2 NVME SSD+4*3.8TB U.2 NVME SSD\8*H100
    299 999,00 US$
  10. Nvidia H200 AI Bundle10, Inspur Nvidia H200 Server, 2x8480+ / 32x64GB / 8xH200

    Condição: Original Novo

    2*8480+\32*64G\2*1.92TB M.2 NVME SSD+4*3.8TB U.2 NVME SSD\8*H200
    308 999,00 US$
  11. H200 AI Server Bundle11, ASUS AI Server, 2×8558/32×64GB RAM/4×3.84TB NVMe/8×H200 GPUs

    Condição: Original Novo

    2*8558\32*64G\4*3.84TB NVME U3\8*H200
    308 999,00 US$
  12. AT600W, Huawei HuaKun AI Inference Server, 32-core Kunpeng 920/128G DDR4/Atlas 300I Duo 96G

    Condição: Original Novo

    Huawei AT600W HuaKun AI Inference Server / 1 × Kunpeng 920 (32 cores) / 128G DDR4 / 1 × 480G M.2 SSD / 1 × 4T HDD / 1 × Atlas 300I Duo (96G) / 4 × GE RJ45 / Single PSU
    7 800,00 US$
  13. AT800, Huawei AT800 Server, 2xKunpeng 920/16x32G DDR4/2xAtlas 300I Duo

    Condição: Original Novo

    Huawei AT800 Model 3000 HuaKun AI Inference Server / 2 × Kunpeng 920 (48 cores) / 16 × 32G DDR4 / 2 × 480G SATA SSD / 2 × 1.92T SATA SSD / 2 × Atlas 300I Duo / 2 × 10GE Dual-Port NIC / Redundant PSU
    25 249,00 US$
  14. AT800, Huawei HuaKun AI Inference Server, 2xKunpeng 920(48 cores)/4xAtlas 300I Duo/16x32G DDR4

    Condição: Original Novo

    Huawei AT800 Model 3000 HuaKun AI Inference Server / 2 × Kunpeng 920 (48 cores) / 16 × 32G DDR4 / 2 × 480G SATA SSD / 2 × 1.92T SATA SSD / 4 × Atlas 300I Duo / 2 × 10GE Dual-Port NIC / Redundant PSU
    22 599,00 US$
  15. AT9508 G3, Huawei HuaKun AI Training Server, 2xKunpeng 920 48-core/16x32G DDR4/4xAtlas 300I Duo

    Condição: Original Novo

    Huawei AT9508 G3 HuaKun AI Training Server / 2 × Kunpeng 920 (48 cores) / 16 × 32G DDR4 / 2 × 480G SATA SSD / 2 × 1.92T SATA SSD / 4 × Atlas 300I Duo / 2 × 10GE Dual-Port NIC / Redundant PSU
    28 989,00 US$
  16. AT9508 G3, Huawei HuaKun AI Training Server, 2xKunpeng 920 (48 cores)/8xAtlas 300I Duo/16x32G DDR4

    Condição: Original Novo

    Huawei AT9508 G3 HuaKun AI Training Server / 2 × Kunpeng 920 (48 cores) / 16 × 32G DDR4 / 2 × 480G SATA SSD / 2 × 1.92T SATA SSD / 8 × Atlas 300I Duo / 2 × 10GE Dual-Port NIC / Redundant PSU
    38 800,00 US$
  17. AT3500 G3, Huawei HuaKun AI Inference Server, 4xKunpeng 920/8xAscend 910B/32x32G DDR4

    Condição: Original Novo

    Huawei AT3500 G3 HuaKun AI Inference Server / 4 × Kunpeng 920 / 32 × 32G DDR4 / 2 × 480G SATA SSD / 2 × 3.84T NVMe SSD / 8 × Ascend 910B 32G / 2 × 25G Dual-Port NIC / Redundant PSU
    105 749,00 US$
  18. AT3500 G3, Huawei HuaKun AI Inference Server, 4xKunpeng 920/8xAscend 910B/32x32G DDR4

    Condição: Original Novo

    Huawei AT3500 G3 HuaKun AI Inference Server / 4 × Kunpeng 920 / 32 × 32G DDR4 / 2 × 480G SATA SSD / 2 × 3.84T NVMe SSD / 8 × Ascend 910B 64G / 2 × 25G Dual-Port NIC / Redundant PSU
    156 199,00 US$
  19. AT600WDD, Huawei AT600W HuaKun AI Inference Server, 1xKunpeng 920/128G DDR4/2xAtlas 300I Duo

    Condição: Original Novo

    Huawei AT600W HuaKun AI Inference Server / 1 × Kunpeng 920 (32 cores) / 128G DDR4 / 1 × 480G M.2 SSD / 1 × 4T HDD / 2 × Atlas 300I Duo (96G) / 2 × 10GE NIC / Single PSU
    10 479,00 US$
  20. R863A, Lead Wealth 2U Rackmount/Tower AI Server, 2x AMD EPYC 4th/5th Gen/Up to 8x NVIDIA L20/24x DDR5

    Condição: Original Novo

    Lead Wealth R863A 2U rack-mount/tower convertible AI server with dual AMD EPYC 4th/5th Gen, up to 8 NVIDIA L20 GPUs and up to 24 DDR5 RDIMMs.

Itens 1-20 de 51

Definir Ordenação Decrescente
Página
Ver como Lista Grelha