The issue of computing power consumption has attracted attention, and green and low-carbon has become a new trend in the transformation of data centers

Interface News

2024-06-30 08:04Interface News official account

The full text is 1578 words, and it takes about 5 minutes to read, so help me highlight the key points

Highlight the point

01With the popularity of AI, computing power, data centers and other topics, the issue of energy consumption has gradually attracted attention, and green data centers have become a new trend.

02The White Paper on the Green Computing Power Index for the Whole Life Cycle of Data Centers was released, aiming to improve the standardization, transparency and recognition of green computing power.

03Experts said that green computing power requires new energy, wind power, photovoltaics, etc. to connect the chain of green computing power operation to promote the sustainable development of green data centers.

04On the other hand, the introduction and promotion of the green computing power index will make the data development chain of the data center clearer and improve the measurement of equipment operation performance indicators.

05To this end, green computing related promotion conferences will be held in various places to promote the deep integration of artificial intelligence and energy.

Technical support is provided by Tencent Hybrid Model

Interface News Reporter |

With the popularity of topics such as AI, computing power, and data centers, the issue of energy consumption has gradually attracted attention.

On June 28, the "Digital Energy Resonance and Green Computing Pilot" Data Center Life Cycle Green Computing Index Forum was held in Beijing, at which the "Data Center Life Cycle Green Computing Index White Paper" was released and the Green Computing Lab Preparation Initiative was launched.

Computing power refers to the ability of computer equipment or data centers to process information, and the construction and demand for computing infrastructure continue to be hot when a new round of technological and industrial transformation is coming. While technologies such as AI, blockchain, and the metaverse are rapidly developing and penetrating into all walks of life, the energy consumption that supports computing power has also attracted attention in the industry. Since the beginning of this year, domestic and foreign business executives and scholars have frequently put forward views such as "artificial intelligence may lead to energy shortages".

Last year, the Ministry of Industry and Information Technology and other six departments jointly issued the "Action Plan for the High-quality Development of Computing Infrastructure", proposing "computing power + energy", which will accelerate the construction of an energy computing power application center, support the intelligent production and scheduling system of energy, and realize the interaction of source, grid and load, multi-energy synergy and complementarity, and intelligent regulation and control of energy demand.

Zheng Weimin, academician of the Chinese Academy of Engineering, said at the above forum that computing power is the foundation of the digital economy, interconnection is the foundation of computing efficiency, and efficient collaboration between computing power and energy is the future of computing power evolution.

According to the White Paper on the Green Computing Power Index of the Data Center Life Cycle, green computing power is a form of computing power for computing infrastructure to achieve green, low-carbon, and sustainable development. The data center is the key foundation of computing power and the physical carrier for the operation of information systems in various industries.

"Green and low-carbon is the trend of data center development, and the infrastructure that supports the production and output of green computing power will become an important carrier to stimulate green investment, and the green and low-carbon development of data centers will not only bring energy-saving and economic benefits to enterprises, but also enhance the image and competitiveness of corporate social responsibility." The white paper said.

Zhou Tianyu, CTO of Hoyinn Data, said in an interview with Jieyin News and other reporters that while the data center is coping with large-scale demand, the first challenge is energy, and if computing power wants to develop, it must first solve the power problem, but it cannot only rely on traditional energy, and new energy wind power and photovoltaic should connect the chain of green computing power operation, that is, computing and electricity synergy, and promote the sustainable development of green data centers.

Second, there are scale limitations, if you want to train a good model, you need a large number of chips in the same physical area, and related technologies need to be broken through.

"The rapid development of artificial intelligence and large models provides the demand for scenario-based green computing power." Zhao Chuhong, vice president of Envision Intelligence, told interface news and other media.

He believes that through the introduction and promotion of the above-mentioned green computing power index, the data development chain of the data center will be clearer, such as the number of machine calculations, how much computing power is, and whether unnecessary loss is minimized in the process, so that the equipment operation results have indicators to measure and traction.

Zhao Chuhong said that the reason why it is called the "green index" is to first see whether the power source of the data center is renewable energy, and secondly, whether the supply side has passed the green power trading and certification, and how to diversify the supply of green energy is an aspect that needs to be considered. At the same time, on the demand side, how to optimize energy consumption and make power coordination more stable through new power systems, source-grid-load-storage or microgrid technology is a part to be considered on the demand side.

At the Data Center Life Cycle Green Computing Index Forum, Yang Xilin, CEO of Hoyin Data, said that the "Data Center Life Cycle Green Computing Power Index White Paper" aims to benchmark green governance achievements into quantitative indexes, improve the standardization, transparency and recognition of green computing power, guide the aggregation of resource elements such as green finance, scientific and technological innovation, and industrial collaboration, and invest in computing infrastructure efficiently, intensively and sustainably.

"The index evaluation system focuses on the safety and efficiency of computing power, green and low-carbon, and management intelligence, and constructs a calculation method covering the whole life cycle of green computing power." Yang Xilin said.

The "Implementation Opinions on In-depth Implementation of the "Eastern Data and Western Computing" Project to Accelerate the Construction of a National Integrated Computing Network" proposes that by the end of 2025, the proportion of green electricity in new data centers in national hub nodes will exceed 80%.

Especially since the beginning of this year, the popularity of "green computing power" has been rising, and green computing power-related promotion conferences have been held in Hohhot, Ulanqab, Qinghai and other places.

Focusing on artificial intelligence and energy, at the summer Davos Forum, Zhang Lei, chairman of Envision Technology Group, also expressed his opinion that artificial intelligence is essentially energy. With the advent of various large models, a "Newton's law" in the field of artificial intelligence has become increasingly clear: intelligence is energy. As long as there is enough energy and enough computing power can be generated, intelligence can be generated.

On the one hand, generative AI relies on powerful data processing capabilities and huge storage capacity, and data centers need to consume a lot of electricity, causing bottlenecks in power supply and power grids. On the other hand, AI plays a key role in building a new power system centered on renewable energy, and can act as a "superagent" to greatly improve energy efficiency.

Zhang Lei believes that the essence of renewable energy such as wind power, photovoltaic and hydropower is weather, and the essence of weather models is statistics. Envision found that artificial intelligence can improve the accuracy of medium- and long-term weather forecasting by 30% in the practice of using energy vertical models to predict weather. In the process of transforming to a new power system, artificial intelligence, the "super brain", can realize the efficient management and coordination of wind power, photovoltaic, energy storage and energy-using facilities, and optimize the management of the future energy system.