XTT9236 AI server
  • XTT9236 AI server
  • XTT9236 AI server
  • XTT9236 AI server

XTT9236 AI server

Brand CHINA

Product origin CHINA

Delivery time Within 15 days

Supply capacity 9999

The XTT9236 is a high-performance GPU server developed on Intel's latest Eagle Stream platform. It supports NVIDIA H20 modules, Intel Gaudi modules, and other GPUs designed for large model training. It is the optimal choice for AI training and large model inference applications.

XTT9236 AI server

Product Features

Leading Architecture Design

  • Intel Xeon Scalable 4th Generation Processors: The platform supports Intel's latest 4th Gen Xeon Scalable processors with a TDP of 350W, ensuring top-tier performance and scalability for demanding workloads.

  • AI Training Module Support: Compatible with the latest AI training modules, including NVIDIA H100 modules, Intel Gaudi modules, and other OAM modules, making it suitable for advanced AI applications.

  • CPU-Switch Connectivity: Features dual interconnect modes to maximize platform performance. This design enhances communication efficiency between CPUs, ensuring optimal data throughput and reduced latency.


Flexible Solution Options

  • Modular Expansion: Supports a range of modules including NVIDIA H20, Intel Gaudi, and other OAM modules, offering flexibility to tailor the system to specific needs.

  • Configurable Storage and I/O: Front-facing storage and I/O configurations can be customized, providing flexibility in storage and connectivity based on user requirements.


Stable Product Quality

  • N+N Redundant Power Supply: The system features an N+N redundant power design for reliable operation, minimizing downtime and ensuring consistent performance.

  • Efficient Thermal Design: Equipped with a high-efficiency cooling system, including 15 high-performance hot-swappable fans with N+1 redundancy, and optional cold plate cooling to maintain stable operation even in high ambient temperatures.

  • Extensive Module Development Experience: Proven experience in developing and debugging multiple module-based products on the same platform, ensuring high reliability and optimized performance.

  • china AI server

Technical Specifications

  • CPU: 2 Intel® Xeon® Scalable 4th Gen processors, TDP 350W

  • Memory: Supports up to 32 DDR5 memory modules, with 8 memory channels per CPU

  • Storage:

    • Supports up to 10 NVMe SSDs

    • Optional support for 2 SATA hard drives

    • 1 onboard M.2 SSD

  • PCIe Expansion:

    • Supports 1 H20 NVLINK module with 8 network slots

    • CPU-Switch X32 interconnect: Supports 1 intelligent network card or 2 OCP 3.0 network cards

    • CPU-Switch X16 interconnect: Supports up to 12 standard PCIe slots

  • Management: AST2600, 1 RJ45 Gigabit management port

  • Basic I/O:

    • 2x USB Type-A

    • 1x Power Button with LED

    • 1x UID Button with LED

    • 1x Status LED

    • 1x Console Port, Type-C

    • 1x Front VGA

  • Cooling:

    • 15 high-performance hot-swappable fans with N+1 redundancy

    • Optional cold plate cooling

  • Temperature: Operating temperature range of 5ºC - 35ºC

  • Power Supply:

    • 2x 12V CRPS power supplies (motherboard) with 1+1 redundancy

    • 6x 54V CRPS power supplies (GPU modules) with 3+3 redundancy

  • Dimensions: 900mm x 447mm x 352.8mm


Why Choose Us

  • Advanced Architecture: Our server leverages the latest Intel Xeon Scalable 4th Gen processors and cutting-edge AI training modules to deliver exceptional performance and scalability for high-demand applications.

  • Flexible and Customizable: With support for various AI training modules and configurable storage and I/O options, our system provides the flexibility to adapt to different use cases and business requirements.

  • Reliable and Efficient: Designed with redundant power supplies and advanced cooling solutions, our server ensures stable operation and reliability even in high-temperature environments. Our extensive experience in module-based product development guarantees top-quality and optimized performance.

  • Proven Performance: Backed by years of development and debugging experience, our platform offers robust performance and reliability, making it a trusted choice for demanding enterprise environments.


Get the latest price? We'll respond as soon as possible(within 12 hours)

Privacy policy

close left right