A few years ago, I began teaching a cloud-computing class at Tsinghua University, considered one of China’s top schools. Kids who get the best scores on China’s national exams typically choose to go to Tsinghua. But I don’t go to Tsinghua once a year just to teach; I also try to learn as much as I can about the state of the art in cloud computing. And China, for several reasons, is a good place to do it. I’ve just returned and, like latter-day Marco Polo, I have a tale to tell.
Unlike the United States and most of the West, where computers were first used by businesses, China’s information age focused on consumer applications. In China, IT did not have to “consumerize”; that’s the way it began. With the country’s late introduction to computers, it was able to skip the floppy-disk era and go straight to applications delivered as cloud services. And while consumer applications led the way, business applications are beginning to emerge — as cloud services, of course.
And as China is the largest growth market on Earth — for any product, but especially for information-powered IT products — it behooves Western executives to understand the Chinese market. Not only are there opportunities, there are also lessons to be learned.
A Cloud Grows in China
The big players in China include Tencent, Alibaba, Baidu, Renren, and Sina, which deliver services similar to Google, Facebook, Amazon, and Twitter. You may be surprised to find out that Tencent, an instant messaging and online gaming service with more than 400 million users, is one of the largest consumer-application cloud services in the world, with a market cap of more than $60 billion.
Two major business-application leaders are Kingdee (cloud management and e-commerce) and Yonyou (cloud enterprise resource planning), and they deliver apps roughly analogous to those provided by Oracle, SAP, NetSuite, Salesforce, and Workday. Large multinationals are probably standardized on Oracle and SAP, particularly for their financials, but these new players are able to tune their applications for the large, Chinese state-owned enterprises (SOEs), as well as for midsize private and public companies.
All these applications use network cloud services. In China these services are provided by the Big 3: China Unicom, China Telecom, and China Mobile. These are extremely large SOEs (China Telecom has more than 300,000 employees). But there is some movement toward deregulation, and a few companies are experimenting with cloud-optimized software-defined networks, implemented across a nationwide backbone, that can simplify network operations by decoupling services from infrastructure.
As is true in many places, the network cloud-service providers also provide data-center cloud services. Along with the Big 3, there are emerging regional and national players such as 21vianet that provide carrier-neutral data in 70 data centers located in 33 cities. It hosts more than 63,000 servers.
The next major step in China (and in America) will be the increasing sophistication of compute and storage cloud services. Aliyun’s service (compute and storage similar to Amazon, Rackspace, or Microsoft) is provided by Alibaba, and it’s in use at such companies as Zoom Interactive, which provides a cross-platform advertising-optimization system. On the pure storage cloud-services front, MeePo, developed at Tsinghua University, is a storage service similar to Dropbox. It is focusing on some enterprise-specific features and relaxing the constraint that you have to have as much storage on your laptop as you do in Dropbox if you want to maintain synchronization. Today, MeePo stores 50 terabytes: not quite as much as Carbonite’s 70 petabytes, but I wouldn’t bet against it getting there.
While these are all examples of more consumer-oriented storage services, NetApp is working with a number of Chinese data-center cloud-service providers to offer more specialized compute and storage cloud services for business.
Just as in America, China is still in the early days of software development cloud services. Recently, 21vianet partnered with Microsoft to deliver Azure, Microsoft’s cloud platform for application development. Google, while available, continues to be partially blocked in China, so its platform for developing and hosting applications on Google’s cloud, AppEngine, has had difficulty gaining traction.
On the operations management cloud-service front, it’s again early days. There are some examples of security management cloud services such as 360 and Jinshan, but both are focused on consumers and small-to-midsize businesses, rather than on large enterprises.
China’s Late Mover Advantage
As a later adopter of technology, China has had the advantage of being able to skip the move from landline telephone to cellular, and therefore its mobile sophistication is ahead of ours. Perhaps the same thing will become true for software as the country skips from licensed-based, self-managed enterprise applications to application cloud services. It’s quite possible that China will also be able to skip the thankless, expensive, and outdated chore of moving computers and storage from loading dock to loading dock and just build all the nation’s new applications on compute and storage cloud services.
The first time I went to China, my friends at Amazon Web Services donated $3,000 worth of compute and storage to my class. I told the students that that could buy the use of a server in Northern California, Ireland, or Virginia for three-and-a-half years. They looked at me blankly, wondering why they should care when they could get a server in Beijing just as easily. But then I explained that that $3,000 equaled the services of 10,000 computers for 30 minutes. No one has ever had access to that amount of computing power for that amount of money. So now the question becomes: What would you do with 10,000 computers for 30 minutes?
The answer to that — when it emerges — may be homegrown in the United States, or it could come from China. Right now, that’s a tossup. But whenever that compute power is put to work, I believe its product will astound the most sophisticated, well-traveled observers, even those who consider themselves, as I sometimes do, latter-day Marco Polos.
Timothy Chou teaches cloud computing at Stanford University. He is the former president of Oracle on Demand and the author of Cloud: Seven Clear Business Models.