Industry News Desk
Cloud Computing: IBM Building its Eighth & Ninth Clouds
North Carolina and Tokyo - with up to three times more computing capacity per square foot than the average data center
Aug. 4, 2008 07:15 AM
IBM said Friday that it is going to build a $360 million state-of-the-art commercial cloud on its campus at Research Triangle Park in North Carolina, intending that the thing be the first built with IBM's New Enterprise Data Center design principles and offer "immense pools of Internet-scale computing technology."
The company is putting another cloud computing center in Tokyo, expecting it to be used by large enterprise customers, universities and government agencies. It said it was the first client-facing cloud in a market as mature as Japan where in-house systems based on mature infrastructures have grown complex and inflexible.
The Tokyo center, the North Carolina center and IBM's seven other cloud centers will be linked so customers can pilot cloud infrastructures and applications globally.
IBM has cloud centers in Dublin, Beijing and Johannesburg. It also offers Blue Cloud, hardware, software and service that let IBM clients offer cloud services to any device anywher
The Research Triangle Park operation is supposed to be quite energy-efficient, a mix of highly dense, real-time water-cooled or air-cooled virtualized servers. In colder months, the data center will switch to free-cooling mode using what IBM called a "water economizer" to cut energy consumption.
The cloud is also supposed to be partially powered by alternative energy to cut carbon dioxide emissions by roughly a million pounds a year. And IBM said the facility's mechanical system is designed to be 50% more efficient than the industry average, saving another 31,799 tons of carbon dioxide emissions a year.
The first phase will involve 60,000 square feet of raised floor space that can be expanded in modular increments. The hardware and software used is supposed to be heterogeneous.
It will supposedly get up to three times more computing capacity per square foot than the average data center.