Ten Trends in Datacenter Facilities

In an attempt to provide the industry a clear picture of the evolution happening in datacenters towards its march to 2025, and to enlighten the way to the future, the globally leading telecommunication solution provider, Huawei has released the 10 trends of data center facilities;

High Density
The CPU and server capacity keep continuously increasing with the evolution of IT computing capacity. As the demand for AI applications increases, the importance of AI computing power further rises. To balance efficiency and costs, data centers will develop towards high density. Currently, the average power capacity in a datacenter is 6 to 8 kW/rack. It is anticipated that the power density of 15 to 20 kW/rack will predominant datacenters by 2025.

Scalable Architecture
Generally, the lifecycle of IT devices is 3 to 5 years, and the power density doubles every 5 years. However, the lifecycle of datacenter infrastructure is 10 to 15 years. The datacenter facility will support the IT device evolution for 2-3 generations. It demands scalable expansion and phased investment for optimal CAPEX in the lifecycle of the datacenter. In addition, the data center must support the hybrid deployment of IT devices with different power densities because of diversified IT services running there.

Green
Currently, the power consumption of datacenters accounts for 3% of the world’s total power consumption. It is estimated that the total power consumption of the datacenter will reach more than 1000TWh by 2025. Energy-saving, emission reduction and operating expense (OPEX) reduction are big challenges. Reducing the power usage effectiveness (PUE) of data centers and building green datacenters are inevitable. It is an irresistible trend to use clean energy and waste heat and to save resources (such as energy, land, water, and materials) throughout the life cycle of the data center. It is estimated that the average PUE of a new data center in China will drop to 1.1 in the next five years.

Quick Deployment
Internet services usually burst in a short period of time, and data and traffic demands on the service side increase sharply. Therefore, datacenters must be rolled out quickly. On the other hand, the data center is changed from a support system to a production system. The faster rollout means faster benefits. The typical TTM of a datacenter is 9 to 12 months, which is expected to be shortened to less than 6 months in the future.

Full Digitalization and AI-enablement
Datacenter facilities will inevitably evolve towards digital and intelligent. With the continuous improvement of IoT and AI technologies, datacenters will gradually evolve from single-domain digitalization in terms of O&M, energy-saving, and operation to full-lifecycle digitalization and automatic driving in terms of planning, construction, O&M, and optimization. AI will be widely applied.

Full Modularization
More datacenters will be constructed in full modular mode to address the problems of slow construction and high initial investment costs. The modular design will evolve from component modularization to architecture modularization and equipment room modularization, finally achieving full modularization of the datacenter. The fully modular design will enable fast deployment, flexible capacity expansion, simple O&M, and high energy efficiency.

Simplified Power Supply Architecture. Lithium Battery Replace Lead-acid Battery
The power supply and distribution system of a traditional datacenter is complex and fragmented, occupies a large footprint, and is difficult to locate faults. A simplified power supply architecture will reduce power conversion times, shorten the power supply distance and footprint, improve the space utilization rate and system energy efficiency. Compared with lead-acid batteries, lithium batteries have advantages in terms of footprint and service life. As the cost of lithium batteries decreases, lithium batteries will be widely used in datacenters in the future.

The convergence of Liquid Cooling and Air Cooling. More Indirect Evaporative Cooling and Less Chilled Water Cooling
GPU and NPU applications generate more high-density scenarios, and liquid cooling systems will become more and more popular. However, some storage and computing services are still in low-density scenarios. To quickly adapt to uncertain IT service requirements in the future, the cooling solution must be compatible with the air cooling system and liquid cooling system. In addition, the complex architecture of the chilled water cooling system hinders quick deployment and easy O&M. An indirect evaporative cooling system, with a modular architecture, will shorten the deployment time and simplifies O&M. In addition, by fully utilizing the natural cooling resources, the power consumption of the cooling system will be greatly reduced. In areas with a suitable climate, the chilled water system will gradually be replaced by an indirect evaporative cooling system.

Dynamic Linkage between Bits and Watts
Reducing PUE doesn’t mean that the overall energy consumption of the datacenter is optimal. Instead of focusing on the data center energy facilities, the energy consumption of the data center needs to be evaluated and optimized as a whole. Through full-stack innovation among facility, IT, chipsets, data, and cloud, bits and watts will work collaboratively to achieve dynamic energy-saving and optimal energy efficiency of the entire system.

Trustworthiness
As the datacenter facility becomes more intelligent, the network security threats will multiply. The datacenter must have six features: resilience, security, privacy, safety, reliability, and availability to prevent attacks and threats from environments and malicious personnel, including network intrusion threats.

Related Articles;

Telstra-Chayora Partnership to Deliver Colocation and Network Services in China

Huawei Datacenter Solution Passes Tolly’s Tests

QTS Selects Ciena Waveserver Ai Platform for Datacenters

TelemaxX Introduces New Ethernet Services with Ciena and Kapsch

The Technologies of Next-Generation Optical Transceivers – PAM4 and 64QAM

Author: Thomas Mathew

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.