[EqualOcean] Overtake with New Infrastructure!

【Yiou-Gong Chenxia】Recently, the announcement was made by the Data Expo, which has been held for five consecutive years:

“In view of the current global COVID-19 pandemic prevention and control situation, to effectively prevent the occurrence of cluster outbreaks, the organizing committee of the Data Expo has decided after careful consideration not to hold the China International Big Data Industry Expo in 2020.

Knowing you from afar, not saying the clouds and sea are deep.”

This is the first time the Data Expo has been canceled.

Hosted by the National Development and Reform Commission, Ministry of Industry and Information Technology, National Internet Information Office, and Guizhou Provincial Government, this global first big data-themed expo, officially upgraded to a national level in 2017, has become a global benchmark for big data development and the most international and authoritative platform for industry results exchange, thanks to its internationalization, professionalism, and market-leading advantages.

At last year's 5th anniversary Data Expo, the organizing committee focused on emerging fields such as big data, cloud computing, artificial intelligence, and blockchain, exporting "Chinese wisdom" and "Chinese solutions" to the global big data community, which is of great significance to the development of big data in Guizhou and China.

The cancellation this time is indeed a helpless move under the pandemic.

However, the exchange and discussion around big data will not stop because of this. Guiyang, the host city of the Data Expo, is China's first national-level big data comprehensive experimental zone, where local construction includes big data centers of companies such as Huawei, Tencent, Alibaba, Apple, and China's three major operators, positioning it as the "City of Data."

Data shows that in 2018, the contribution rate of big data to Guizhou's economic growth has exceeded 20%. The "China Digital Economy Development and Employment White Paper (2019)" recently released by the China Academy of Information and Communications Technology shows that in 2018, Guizhou's digital economy growth rate and digital economy employment absorption growth rate both ranked first in the country.

With the introduction of the "new infrastructure" national policy, known as a market rescue, big data centers have become a high-frequency term. How do they differ from traditional data centers? What role do they play in the overall push of new infrastructure?

Recently, we chatted with several industry experts in the field of big data, including Zhao Wei, Senior Vice President of, Peng Zhen, Senior Vice President of Inspur Group, Gao Junjie, Director of Transwarp Technology, and Yu Haizhong, CEO of Ruifan Technology, about the matters concerning big data centers under the new infrastructure.

Looking at history: Big data centers are no longer the previous IDC

Guiyang was once called "Foxconn in the cave."

Information shows that Foxconn in Guiyang is not only a foundry for Huawei, Apple, and Xiaomi in Guizhou but also has data centers and R&D centers, even hollowing out three mountains in Guiyang to install tens of thousands of servers inside.

Does the construction of data centers require such a large-scale effort?

First, we need to understand what a data center is.

It refers to facilities used to house computer systems and related components, such as telecommunications and storage systems. It generally includes redundant and backup power supplies, redundant data communication connections, environmental controls (such as air conditioning, fire extinguishers), and various security devices, combining real estate attributes and IT attributes. Moreover, large data centers consume electricity about the same as the industrial business scale of a small town.

The "Guizhou Province Big Data Industry Development Application Plan Outline (2014-2020)" also mentions: Guizhou Province has a subtropical monsoon humid climate, with an average summer temperature below 25°C; electricity prices have a competitive advantage, and the average industrial electricity price in Guizhou Province is significantly lower than in other domestic regions.

Therefore, the location of data centers is often in areas with vast land resources, conducive to server construction, abundant power resources, and lower electricity prices, such as Guiyang in Guizhou and Ulanqab in Inner Mongolia.

This is a very obvious traditional data center construction, namely the IT infrastructure of big data, biased towards hardware servers.

With the development of the internet, hardware servers access bandwidth, high-performance local networks, have a safe and reliable machine room environment, professional management, and a complete application service platform, on this basis, they can provide internet infrastructure services (server hosting, virtual hosting, email caching, virtual email, etc.) and various value-added services (site rental services, domain name system services, load balancing systems, database systems, data backup services, etc.).

At this time, data centers are called IDC, namely Internet Data Center.

In recent years, a new type of data center based on cloud computing architecture, highly virtualized, automated, and energy-saving has emerged and gradually expanded in scale, which is now commonly referred to as cloud computing data center. Its virtualization of servers, storage, networks, applications, etc., allows users to call various resources on demand, manage related businesses in a process-oriented automatic manner, and meet green energy-saving standards in all aspects.

Unlike traditional data centers and IT systems that are separated, cloud computing data centers emphasize more on collaborative optimization with IT systems and more on the reliability and safety of machine rooms, there are significant differences between the two in terms of operational efficiency, service types, resource allocation, and charging models.

Guizhou Province once stipulated: Except for special needs, all provincial government departments will no longer purchase servers, switches, storage, and other hardware equipment on their own, nor will they build their own machine rooms, government data will be uniformly stored in the cloud.

Going to the cloud is a trend. Because, the cost of the cloud is lower, and the efficiency is higher.

So what is the big data center in the new infrastructure this time? How does it differ from IDC and cloud computing data centers?

According to the interpretation of the National Development and Reform Commission, new infrastructure mainly includes three aspects: First is information infrastructure. Second is integrated infrastructure, third is innovative infrastructure. The latest official definition frames big data centers within the computing power infrastructure of information infrastructure.

Several experts told Yiou that the big data center in the new infrastructure this time will definitely be combined with domestic production and the overall promotion of domestic emerging industry upgrades.

In the future, it will not just be IDC construction.

The implication is that big data centers are no longer simple facility provision bases for building many machine rooms and networking to provide services, but through this layer of infrastructure to serve upper-level applications and business services, providing computing power foundation for the digital industry upgrade of various industries in China.

Investment involving big data centers will be the most intensive investment in the new infrastructure, and it will become a node in various links of China's industrial chain.

"Big data centers must have new businesses, which are new growth points, these new growth points will not cause urban and population polarization like real estate, but will make the country's overall strength more balanced due to the attributes of infrastructure."

Looking at the present: Domestic substitution, who is entering the big data center?

There is national policy above, and market demand below. Perhaps some speculators want to enter this trend again.

But in the wave of big data centers, temporary radicalism and opportunism will not make any splash. There are also long-term layouts of major giants and the long-cycle investment return attributes of infrastructure. Yu Haizhong said that for enterprises, the value gained may not only be economic value but also greater social responsibility value.

In fact, in the industrial chain of big data centers, it can be subdivided into many categories. The industry has also listed big data IT infrastructure, data resource acquisition, big data storage, data integration and analysis, data security, data application and consumption as specific divisions.

We can summarize the above categories into two types, one is the big data hardware IT infrastructure that requires actual site and equipment maintenance and operation, and the other is the software architecture.

Among them, the software architecture includes three layers:

1) Big data technology platform: including big data storage (database), and the data integration and analysis platform formed by using the open-source Hadoop distributed storage database.

2) Big data middle platform: plays a role in integration and sharing, also includes data security.

3) Business application layer: refers to the upper-level data application and consumption.

In this contact with big data service providers, comparing them, has its own commercial database, mainly providing underlying structured data services; Transwarp is a closed-loop data platform detached from Hadoop (distributed system infrastructure), and has one-stop data services on this basis; Ruifan is a data middle platform based on Hadoop, which can cooperate with other service providers; Inspur not only has software services but also its own hardware. These manufacturers are leaders in the field of big data centers, but their positioning is different, and their strategies for future big data centers are also different.

But they are all doing the same thing, namely going IOE (i.e., IBM's minicomputers, Oracle databases, EMC storage devices) and domestic substitution.

As Zhao Wei said, this time the country has included big data centers in the new infrastructure category, which is an important market opportunity for the overall shift of the domestic information industry to a domestic ecological system. The previous practice of purchasing foreign products for information system construction will undergo significant changes, and new infrastructure projects will introduce more domestic products.

Taking the core technology MCU chip as an example, domestic MCU, whether in market share or technological advancement, cannot compare with foreign companies. For Chinese companies, the mainstream market currently occupied is still in the 8-bit MCU, accounting for about 50%. The 16/32-bit MCU accounts for about 20% each. This means that domestic MCU application fields are mostly concentrated in low-end electronic products, while the mid-to-high-end electronic product market is still in the hands of foreign companies.

In recent years, autonomous and controllable big data software and hardware have received encouragement and support from national policies, the government has strengthened the emphasis on intellectual property rights, strengthened the emphasis on chips, original databases, software, and artificial intelligence products, and encouraged the development of original technology enterprises. This not only promotes the establishment of a good market atmosphere but also gives domestic manufacturers opportunities.

On the one hand, domestic substitution is giving domestic enterprises more choices, breaking the monopoly of expensive foreign products, providing enterprises with domestic products that are not inferior to foreign products in terms of performance, stability, and reliability, allowing enterprises to choose products calmly in the construction of big data centers. On the other hand, at the national level, autonomous and controllable core technology also involves national security and national status.

In the current situation, whether it is the sanctions of the Sino-US trade war or the risk of reverse globalization under the pandemic, having core technological capabilities means having the most hardcore voice.

The launch of new infrastructure this time, is a market rescue and also a curve overtaking.

Yu Haizhong believes that besides promoting employment under the pandemic, China's industrial structure has been optimizing towards high-end industries, which has also contributed to the booming development of the domestic big data and artificial intelligence industry. The implementation of new infrastructure is laying the foundation for these high-end industries, after the new infrastructure is perfected, China's industrial structure will rise to a higher level, which will lead to a wave of big data construction demand.

And the old infrastructure is currently in a relatively saturated state. China's railways, highways, bridges, and other infrastructure have already far exceeded those of European and American countries in total. For example, China's high-speed rail mileage exceeds the total of all overseas countries. Currently, the cost-effectiveness of investing heavily in old infrastructure is not high, and proposing new infrastructure conforms to the development of the times.

For China, this is an opportunity.

Facing the pandemic, the US's internal aid plan basically does not include parts related to new infrastructure industries, which means that China is accelerating, while the US is slowing down, especially in 5G construction.

According to industry analysis reports, the number of 5G base stations built in the US is currently only around 10,000 to 15,000, while the number of 5G base stations in China has reached 130,000, and it is expected to reach 600,000 by the end of 2020.

In addition, the US's rail transportation is also weak, and there is no ultra-high voltage grid; European countries have even fewer with corresponding technological development capabilities in these fields.

Although there is still a certain gap between China and Western countries in terms of technological strength, the leading position in some technological application fields such as "New Four Great Inventions" (high-speed rail, mobile payment, shared bicycles, and online shopping) also proves our potential.

This reflects that the degree of IT and data construction in China is already ahead of foreign countries. The development of new infrastructure this time will greatly expand this advantage.

We can even predict that in the past, Chinese companies purchased foreign technology, now China's big data, artificial intelligence, 5G, high-speed rail, and other technologies will be exported overseas, achieving curve overtaking.

Looking at the future: What is the blueprint for big data centers?

For the government as the leader and enterprises as participants, the core of new infrastructure is no longer internet construction, but intelligence and the interconnection of everything, so the cost will be higher. If this layout is to be done well, further refinement and implementation are needed from talent to investment attraction.

From the perspective of difficulties, infrastructure investment is undoubtedly a huge cost. Therefore, in terms of policy guidance and support policies, measures such as subsidies, loans, and taxes should be introduced with corresponding regulatory policies, just like new energy subsidies. Moreover, the construction of big data centers, due to its industrial linkage, will also bring broader effects.

Peng Zhen believes that data centers are the existence form of computing power, the national new infrastructure policy will drive further investment and development of data center infrastructure, and the investment in computing equipment such as servers and storage, as well as cloud computing software and other platform software, is expected to increase, which is a major benefit for the entire industry.

Specifically, compared to traditional data centers and technology, cloud data centers and related products, edge data centers and related products, will achieve better growth.

As of the third quarter of 2019, there are 504 super-large-scale data centers worldwide, these data centers, according to the most conservative estimate—each accommodating 50,000 servers, can accommodate more than 25 million servers in total, while the annual global server sales are less than 13 million units, and there are 151 super-large-scale data centers planned or under construction, which means most of the servers sold in the future market will be deployed in super-large-scale data centers.

According to IDC data, in 2019, global enterprise and government investment in cloud infrastructure officially exceeded spending on traditional non-cloud IT infrastructure, the growth of the global IT infrastructure market in recent years mainly comes from innovative applications, traditional IT infrastructure has not grown, and even shrunk. The national new infrastructure investment will also maintain this characteristic, mainly used for the procurement of innovative products such as cloud servers, accelerating the adjustment of the market structure of IT infrastructure such as servers, and realizing industrial upgrading.

On the other hand, edge computing will develop faster. 5G is about to enter large-scale commercial use, which means it will also enter a rapid development stage. Gartner predicts that by 2020, 50% of large enterprises will include edge computing in their plans, more than 50% of industrial IoT analysis will be conducted at the edge, up from less than 10% in 2017; by 2022, more than 50% of enterprise data will be generated or processed outside of data centers or the cloud.

Overall, data centers are the engine of all new infrastructure. The country has launched seven major industry directions for new infrastructure, supporting more comprehensive data collection by big data centers through 5G base station construction, with big data centers driving and empowering artificial intelligence, applying artificial intelligence technology to industrial internet and new energy vehicles. At the same time, supporting the construction of ultra-high voltage, intercity, and urban rail transportation through big data centers.

The importance and pioneering nature of big data centers are self-evident.

Epilogue

The latest news, on May 15, 2020, the US Department of Commerce website appeared an announcement from the Bureau of Industry and Security (BIS) of the Department of Commerce, announcing changes to the Export Administration Regulations (EAR), the gist of which is to completely cut off Huawei's global wafer foundry sources.

In response, the "Global Times" reported that sources said that if the US side ultimately implements the above plan, China will strongly counteract to safeguard its legitimate rights and interests.

Domestic media commented: Huawei will not be the only victim under the pressure of US technological hegemony, nor will it be the last.

Where will China go under new infrastructure? Only with technological innovation can the country rely on it to be strong.