With over 384 million people on the internet, 900 million mobile phones and 94.3 % of Chinese internet users having broadband access, why is the adoption of Cloud Computing slower in China than the rest of the world? Key driving factors for cloud adoption in US and elsewhere such as efficient cost structure, flexibility and ease of use, access to technology and solutions which were earlier not affordable don’t seem to be driving demand for cloud in China as much.
Cloud make a lot of sense for start-ups and they have been chiefly instrumental in driving SaaS and IaaS adoption so far. If you look at the US Venture Capital figures (2010) of $21.8 billion vs. that in China – ~$3billion, there is hardly that kind of start-up-led cloud adoption push in the context of China. Interestingly, small and medium businesses are the backbone of economy not only in China but in APAC. Their primary interest in cloud is in its capability to handle unpredictable workloads and not justs costs or accessibility. Remember, cost structure of internet services is drastically different from that in the U.S. and other parts of the world. For example, cost of bandwidth is significantly higher but the market price for a virtual machine is lower than that in the US. There is cheap skilled manpower available and alternative avenues of hardware access in the form of rental/leasing facilities which can substitute for small scale IaaS-like requirements with similar pay as you go model and the rental company maintaining and upgrading the hardware.
In addition to the difference in start-up culture, Chinese data centers may not be going through similar transition after all, as their US counterparts are. There are many older legacy laden data centers in US that are due for technology and facility overhaul to beat the single digit utilization rates and higher power consumption and PUE metrics that result in wasted capacity and higher overall operating costs. In the US context, such data centers benefit tremendously from the virtualization and cloud computing paradigms whereas the Chinese ones are relatively newer or probably fewer legacy data centers than in the west. If you draw parallels from the High Performance Computing (HPC) world - China has recently upped its ante with World’s #1 Tianhe and push towards being a global supercomputing power. The total number of indigenous supercomputing systems in China is higher than in any other country. This illustrates China’s broader strategy for developing its own supercomputing infrastructure through indigenous system expertise and capability without relying on imports poentially down to the CPU level. Godson (aka Loongson) is a MIPS-based processor family, developed by the Chinese government-backed Institute of Computing Technology (ICT) in the Chinese Academy of Sciences. In 2007, a supercomputer with the name of KD-50-I was constructed, using 336 Godson-2F processors to deliver one teraflop of performance.
Similar to HPC domain, Cloud Computing in China too has the same indigenous streak. Most of the foreign players are wary of Chinese government rules and regulations and so are the customers in China. They are more comfortable using local service providers as opposed to external players – take the case of qq.com which hit Microsoft IIS and Apache market and gulped 12% of Webserver market share in 3 months!
After government regulations and the extra administrative overheads that it entails, it is the data security in clouds that Chinese organizations are wary of . They appear to be paranoid regarding having their data out of China even more than the US cloud adopters! An Accenture survey suggests that Chinese organizations are averse to entrusting foreign providers with their data. Less than half say they are willing, even if the provider keeps the data within China’s borders. The number falls to below 20 percent for foreign providers that do not have cloud data centers in China.
Absence of clarity on data privacy laws in China is another hurdle to cloud adoption. There is no doubt that cloud computing is viewed as a beneficial IT paradigm in general but the compelling reasons for mass adoption are still fuzzy when it comes to Chinese landscape. If you look at virtualization adoption – one of the cornerstones of cloud computing – China lags in virtualization adoption as well. Once the Chinese broadband speeds get better and match up with the other developing countries and basic cloud computing concerns of Chinese organizations are addressed especially reliability of cloud services and the sensitivity of their data we may see a spurt in cloud adoption. As a culture, Chinese organizations are reluctant to use foreign cloud providers or turn to start-up companies, so there may be local players (Kingdee, Baihui, Ufida) that pop-up higher in cloud arena in near future. Boundaries in clouds are fuzzy (literally and also in cloud computing) and requirements such as ‘non-floating’ data location, privacy could end up tethering data in clouds or creating continental/country-private clouds!
