top of page
資料中心

Year of writing: 2020
Author: Tony

Related stocks:  Amazon , Google , Microsoft , Baidu , Tencent (HKG: 0700) , Lotte (TYO: 4755) , Hon Hai (2354) , Inventec (2356) , Quanta (2382) , Wisdom (6669) , ASRock (3515) , Mitac (3706) , Facebook, Adobe, Apple, PayPal, Salesforce, eBay, Zoom, Twitter, Line (TYO: 3938), Verizon, AT&T, China Mobile (HKG: 0941), Spotify, Netflix

Data Center Data Cente r  Part 2

In-depth analysis of cloud application trends

There are hundreds of cloud applications, from renting cloud space, AI computing performance to applications around life, but which ones have high prospects? Which are the areas where the white-label server warriors must compete? Which ones have strong demand for servers? Which countries will win again? This article will take readers to understand the prospects of future applications from the application level.

Competition between the U.S. and China

Generally, white-brand server vendors divide cloud customers into first-tier cloud companies (Hyperscale Data Center, the inner circle below), and second-tier cloud companies (ENTERPRISE DATA CENTER, the outer circle below), although The second-tier customers are usually not as large as the first-tier customers (Google, Amazon, Microsoft, Facebook, Baidu, Alibaba, Tencent), but the large number of applications is different, and it is enough to provide enough nutrients for the white-label server to grow. The following figure divides the big cloud companies into the world's two major powers into Google, Amazon, Microsoft, and Facebook in the United States (on the right of the inner circle), and BAT Baidu, Alibaba, and Tencent in China (on the left of the inner circle). Although most Taiwanese individuals and business users are familiar with American Pie cloud services, China has a domestic market and collects a large amount of offline data. Compared with American companies which usually only have online data, China needs more servers in China. In addition, according to Gartner research, Southeast Asia is the fastest-growing cloud market, making Southeast Asia a battleground for the military. Major Chinese factories are also accelerating investment in Southeast Asia, and Alibaba Cloud has become the top three cloud providers. . Therefore, the Chinese cloud market cannot be ignored.

雲端服務企業(資料中心)

The outer layer of the above figure divides the cloud service SaaS into several major applications, including e-commerce and commercial software, communications and communities, entertainment and games, and telecommunications services. "E-commerce and commercial software" will have more new applications launched in the future, and more and more companies that provided offline software downloads in the past will tend to provide online services to maintain a sustainable cash flow. In the pursuit of smoother video calls, "Communications and Community" has also transmitted a huge amount of data. After the new crown virus, video remote work has become increasingly popular. The fast-growing conference software Zoom currently uses AWS, Azure and Oracle It’s worth paying attention to whether it is possible to build a data center in the future and which supply chains can get orders. "Entertainment and games" Because of the huge amount of data that needs to be processed, the server and the demand for data are relatively high. The audio-visual streaming platform Netflix has also begun to build its own data center. The server orders are won by ASRock, and many technology companies have launched them. Cloud game streaming, including Google launching Stadia, Microsoft launching Xcloud, in order to promote players to have a better real-time experience, game streaming data center will need more top-level, more equipment, this is also in a variety of data center applications, I think the development opportunities are the most impressive. "Telecom Services" is like  Part 1  According to the introduction, in the past, the telecom data center was mainly controlled by the three major telecom equipment vendors, and others were not allowed to enter. But now white-brand equipment vendors can already use cloud technology to enter the telecom field, which is a brand new field for server vendors.

Comparison of the three major cloud data centers

三大雲端資料中心比較

Although China's power cannot be ignored, in recent years, the continuous suppression in the United States and the gradual saturation of China's internal cloud market, the United States is still the main customer of Taiwanese manufacturers. We compare the world's three largest cloud data centers. Amazon has the most data centers and also has the market share. However, the number of cloud centers is not necessarily proportional to the amount of data that can be processed. It still depends on the amount of equipment in the cloud center. Some may only have more than 10,000 servers, and some may reach 100,000. 10 times. In the past, Microsoft had a leading position in the commercial market for a long time. Therefore, Microsoft also has its advantages for cloud services, which are mainly commercial services. Google has invested heavily in data centers in recent years, and capital expenditure is the highest of the three. It is expected to become the second largest in the market. The best thing Taiwanese know is that it has a Google data center in Changhua and is about to build a second in Tainan. Among the current 38 seats, only nearly half are leased out, and the rest are for Google’s own use. Google’s advantage lies in its innovative technology. The powerful AI allows users to use it easily without having to train basic models by themselves. I think Google Cloud will be the most explosive.

Most of the data centers of China’s three major cloud vendors BAT (Baidu, Alibaba, Tencent) still use branded servers, and the servers are mainly from local Chinese vendors, Inspur, Huawei and other server vendors. However, Taiwan brand servers can also be used indirectly. Factories benefited, especially Inspur’s main foundry Mitac.

Amazon 資料中心 AWS 分布

Amazon data center AWS distribution

微軟資料中心 Azure 分布

Microsoft Data Center Azure Distribution

Google 資料中心 GCP 分布

Google Data Center GCP Distribution

bottom of page