The partnership seeks to capitalize on the growing interest in so-called “cloud computing,” idgml-ce941b51-f87e-44eb-b977-fca042e98da0 where companies run applications in remote data centers that are managed and maintained by another service provider.
It can also be used for general application development and testing, said Richard Payling, Capgemini vice president of global outsourcing.
“All you do is squirt applications to the cloud,” Payling said.
Applications will be hosted within Amazon’s Web Services, the company’s network of computer, storage and database services centered around Amazon’s Elastic Computer Cloud (EC2). EC2 is essentially a network of data centers where companies can rent servers to run their own applications.
Capgemini is targeting companies that want to deploy Web-based applications quickly, Payling said. Application development and deployment is often slowed down by the procurement phase, where companies have to buy their own hardware.
Enterprises will save money by renting Amazon’s infrastructure instead, Payling said. The arrangement also scales faster, as companies can just rent more servers as their computing demands rise.
Since September, Capgemini has conducted a pilot program with three clients, one of which is a vehicle manufacturer. The company used Amazon’s infrastructure to host its SharePoint application, which it used as a knowledge portal, said Andrew Gough, Capgemini’s global lead for cloud services.
Capgemini is expecting companies will increasingly move toward cloud computing as they gain confidence in the capabilities of the concept, Gough said.
Now, companies tend to be more receptive to moving noncritical applications to a cloud service rather than ones which if they failed would shut down the business, he said.
Capgemini and Amazon will offer a 99.95 percent uptime guarantee a year. That works out to no more than four hours and 30 minutes or so of downtime per year.
Capgemini will bill customers for its consulting services and Amazon’s infrastructure. Use of Amazon’s infrastructure is billed per CPU (central processing unit) used per hour, with data billed by the megabyte.
Companies are looking for ways to reduce their IT spend now that the global economy has taken a turn for the worse, Payling said. “Everyone is looking for ways of doing more for less,” he said.