Ai Compute | KAYTUS

AI Compute Pod 200

Intelligent
Computing Power Pod

A modular infrastructure integrating AI, computing, storage, and networking.
Designed for high-density deployment, energy efficiency, security intelligence, and
rapid delivery, it provides a solid foundation for next-generation AI applications.

Dowanload datasheet

Product Overview

The AI Compute Pod 200 Intelligent Computing Power Pod is KAYTUS’s next-generation modular data center product. Integrating AI, computing, storage, and networking, it delivers a complete infrastructure solution for high-density computing scenarios.

Standardization
Engineering-Free Deployment
Optimized Operational Efficiency
Intelligent Management

Wide Range of Application Scenarios

Autonomous Driving

Biopharmaceuticals

AIGC

Four Core Competencies

High Density: Industry-Leading Power Density

Cooling Max: 50kW/rack

Liquid Cooling Max: 240kW/rack

Supports ultra-high-density server

Network routing shortened by 50%

Minimalist links, 30% space saving in power distribution
Green & Energy Saving: PUE as Low as 1.1

Unique enclosed hot and cold aisle design with AI intelligent control, cooling supply frequency-matched with IT load in seconds.

Enhances heat exchange efficiency, "Zero" noise

Reduces PUE from 1.5 to 1.3 within 1-2 weeks

Compatible with general/AI/HPC liquid cooling servers
Fast Delivery: Construction Cycle Shortened by 70%

Component-based modular shipping, plug-and-play, eliminating traditional aisle enclosure setup.

Supports 350kW IT equipment cooling

N+1/N+2 redundancy for flexible configuration

All-modular rapid assembly method

Security & AI

Smart & Secure: Original Dual-Group Control

Achieve comprehensive security management through 3D scene reconstruction, power link monitoring, and thermal mapping. Triple leak detection ensures safety in high-density liquid cooling environments.

Eliminate local hot spots

N+2 Air Conditioning Redundancy

Triple Leak Detection

Technical Specifications

Physical & Environmental Parameters

External Dimensions: L≤15000mm; W=1800/2000mm; H=2800/3000mm

Reliability Design: Supports dual power feed, UPS N+1/2N architecture.

Rack Configuration: 42U/48U (W600/800mm, D1200/1400mm)

Application Scenarios: Suitable for large, medium, and small-scale high-density micro-modular data centers, supporting AI and HPC platform construction
Cooling Systems

300mm In-row Air Cooling25kW

600mm In-row Air Cooling50kW

600mm Liquid Cooling End60kW

High Performance Liquid Rack240kW

Features: Full frequency conversion, low-load dehumidification, refrigerant leak detection, online fan maintenance

Liquid Cooling Model Support

HPC Liquid Cooling 72 nodes/rack (144 CPUs), 130kW | i24, 5180/5280

AI Server Liquid Cooling 48 GPUs/rack, 48PFlops, 80kW | 5688/5698, 54X8
Power Systems Specifications

33kW /
4-8 units High Power PDU

350kW /
2-8 units Smart Powershelf

33kW /
6-8 units Smart Busbar

Smart Management Platform

3D Scene Reconstruction

Based on digital twin technology, 1:1 reconstruction of micro-module layouts. Covers secondary piping, racks, and manifolds with precise component positioning.
Flow Control & Thermal Mapping

Flow adjusts in real-time with chip power consumption for precise cooling. Records chip temperature curves and provides alerts for long-term overheating.
AI Optimization

12-second interval temperature updates with AI modeling to predict heat load trends. Dynamically adjusts cooling output 12-48 seconds in advance.

12s

INTERVAL

48s

PREDICTION

Get a Customized Intelligent Computing Power Pod Solution

Contact our team of experts today to create an efficient, stable, and scalable intelligent computing infrastructure tailored for you.

Contact usDowanload datasheet