MorganStanley predicts that by the end of 2012, Cloud Computing- including Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS), and Software-as-a-Service (SaaS) will be a US$400 billion dollar global industry- a fantastic rise for such a new technology. But behind the rapid rise of the cloud lies one nagging question- do we have the server space to handle it?

This is a big question, as Cloud applications are more data center intensive than almost any other technology to emerge before, and the total consequences of the shift from locally hosted software and platforms to vast server farms have yet to be fully assessed. Several trends are quite clear, however:

-Higher Utilization Rates: Public cloud architecture and server virtualization mean that server workloads can increasingly be handled utilizing resources on underutilized servers, thus cutting down overall demand for server hardware. While the explosive growth in overall resource demand will still drive increased server sales, cloud architecture will allow those systems to be utilized in ever more efficient ways.

-Changes in Location Utilization: Virtualization won’t just be for public cloud or managed servers anymore. Increasingly, companies using on-site data centers with multiple servers are using virtualization to create private cloud systems, upping their ability to utilize existing hardware.

-Workloads Moving to More Efficient Environments: In 2005, less than five percent of server workloads were handled by systems under managed hosting, and barely one percent of workloads were on then-nascent public cloud systems. Morgan Stanley suggests that by 2014, public cloud systems may handle as much as 25 percent of workloads, and managed hosting systems could take up another 15 percent, indicating that the vast majority of server work will be ported to higher-utilization environments.

Thus, while overall server demands are definitely set to rise, they might not rise as much as one might think given the vast needs projected by many companies (such as Huawei, whose projected server demands just to meet the needs of their cloud computing phone handsets represent nearly 0.5 percent of the entire capacity of the global internet).

Some of the companies best poised to meet the server demands of cloud computing are right here in the sinosphere- in Taiwan, to be specific. Quanta Computer and Wistron have emerged as two of the most significant technology leaders in the server industry, and are well positioned to take advantage of the Asian cloud boom.

What are the roots of their advantage? The shift in power from traditional OEM companies, like HP, to ODM companies- contract manufacturers who can ship off-brand product directly to buyers. Over the past two decades, Taiwanese contract manufacturers have developed their in-house engineering and design staff to the point that they can develop products entirely independent of overseas engineering firms in the US, EU or Japan- and can now pass on the savings to customers, especially customers in greater China. With the boom in server demand in China from ambitious cloud hardware makers, it seems likely that Quanta and Wistron will go from back-end manufacturers to front-end brands among Chinese data center operators.

This is just one more instance of the disruptive power of the cloud- and Silicon Valley had better be on notice, or they could completely miss out on this aspect of the revolution.