Chindata Group Unveils AI Data Center Total Solution 2.0
Press Releases
Oct 25, 2024
BEIJING, Oct. 25, 2024 /PRNewswire/ –Chindata Group officially unveiled its latest innovation, the AI Data Center Total Solution 2.0, at the 2024 China Computing Power Conference. Designed to meet the growing demands of artificial intelligence (AI) workloads, the solution addresses critical challenges such as high-density cabinets, hyperscale facilities, and heterogeneous computing environments.
The launch took place during the Data Center Technology Innovation and Transformation Forum, co-hosted by Chindata Group and the China Academy of Information and Communications Technology (CAICT). Held at the same conference, the forum brought together industry leaders from CAICT, China Telecom, Intel, Baidu, Alibaba Cloud, Inspur, Seagate Technology, and others to explore the latest advancements in intelligent computing and IT infrastructure for the AI era.
Addressing Evolving Demands in AI Workloads
The sharp rise in power consumption among GPUs and servers has led to data center clusters scaling beyond hundreds of megawatts. These trends present new challenges for traditional data center operations. As a leading data center operator, Chindata Group has developed a proactive strategy to address the increasing demand for computing power.
The AI Data Center Total Solution 2.0 incorporates advanced prefabrication and modular building models, the innovative “X-Power” power system, the “X-Cooling” cooling solutions, and an AI-driven comprehensive operation and maintenance platform. This state-of-the-art solution is designed to support the accelerated growth of artificial intelligence and ensure high-quality and flexible business operations for customers.
Key Features of AI Data Center Total Solution 2.0
1. Flexible Prefabrication and Modular Construction: Prefabrication and modular construction models are capable of being arranged and manufactured in various stages, which allows for the agile deployment of business operations while accommodating a wide range of hardware configurations, including CPUs and different types of AI accelerators. This modularity improves IT performance per square unit and addresses the growing high-density requirements of AI technologies. This solution delivers a fast, efficient response to the need for high computing power in the era of large models and general artificial intelligence (AGI). Utilizing this construction methodology, a hyperscale project can be completed and handed over to the client within 6 months in China, and within 8 months across the rest of the Asia-Pacific region.
2. X-Power Modular Power System: The upgraded “X-Power” system integrates over ten traditional components of power distribution into six primary modules: substations, generators, medium voltage units, low voltage units, batteries, and IT support equipments. These modules can be deployed in phases according to specific needs, allowing for flexible and scalable expansion. By reducing complexity, this system enhances power efficiency and accelerates the delivery of hyperscale facilities while maintaining the highest quality standards. The X-Power system is capable of supporting server cabinets with power densities of up to 132kW, providing robust power capacity for the most demanding AI and high-performance computing.
3. X-Cooling Solutions: The “X-Cooling” solutions offer comprehensive cooling systems that combine air cooling, liquid cooling, and hybrid cooling methods to address the high-density demands of AI-ready data centers. With capabilities ranging from cabinet power densities of 8kW to 150kW, and achieving a PUE as low as 1.10 with immersive liquid cooling, this solution integrates multiple cooling technologies and products, adapting to architectural, environmental, and geographical constraints.
As traditional air cooling methods reach their limits in high-density environments, there is an increasing demand for liquid cooling in hyperscale data centers. In response, Chindata has initiated the development of industry standards for cold plate liquid cooling systems in China, leveraging the experience gained from developing and deploying X-Cooling solutions. These standards will ensure quality control throughout the lifecycle of the systems.
4. AI-Driven Operations and Maintenance: Chindata’s integrated monitoring platform, combined with its data center operational management system, enables more automated and standardized management processes. By incorporating AI-driven solutions, the platform optimizes the control of critical systems, particularly the cooling systems, ensuring maximum efficiency and reliability. Chindata continues to refine its operation and maintenance protocols to meet the evolving demands of AI data centers.
Industry Leadership in AI Infrastructure
Zhang Binghua, Chief Technology Officer at Chindata Group, commented on the launch:
“The rise of AI is fundamentally transforming the landscape of data center technologies. Chindata has always been committed to leading technological innovation, focusing on modularity, integration, and advanced prefabrication. With the AI Data Center Total Solution 2.0, we are well-positioned to meet the diverse computing needs of the AI era, ensuring the efficient translation of electrical energy into computational power.“
About Chindata Group
Chindata Group is a leading carrier-neutral hyperscale data center solutions provider in the Asia-Pacific region, pioneering next-generation hyperscale data centers in China, India, and Southeast Asia. The company offers end-to-end services across the facility lifecycle, including planning, investment, design, construction, and operation. With a robust international development team, Chindata is expanding its data center clusters in Malaysia, India, and Thailand, and seeking further business opportunities in emerging markets across Asia-Pacific.
For media inquiries, please contact: media@Chindatagroup.com
View original content:https://www.prnewswire.com/news-releases/chindata-group-unveils-ai-data-center-total-solution-2-0–302287161.html
SOURCE Chindata Group