VMware Inc. announced a new program that will allow its service-provider partners to offer pay-as-you go computing services similar to those from Amazon Web Services (AWS).
Using VMware's vCloud virtualization software, managed service providers such as Terremark and Verizon Business will allow customers to configure virtual servers over the Web and pay for the computing capacity they use on an hourly basis with a credit card, much as AWS does today.
VMware CEO Paul Maritz demonstrated the services, which will be branded vCloud Express, in his keynote at the start of VMworld today. The company will hold a press conference with service providers later today to provide more details.
VMware appears to be pitching the services as ones that will support a wider choice of OS environments than AWS does, as well as offer lower starting prices.
“When you see the vCloud Express logo you know it means fast and cheap — or rather I should say, fast and cost-effective,” Maritz said in his keynote.
The services are “aimed at the end of the market where you want very rapid provisioning,” he said. Maritz highlighted test and development work as one possible use case, though some service providers will offer service level agreements to support production applications, VMware said.
Terremark, the managed hosting company that VMware invested in earlier this year, has already posted some details and pricing for its services on its Web site. Savvis and AT&T will also offer the services from their data centers.
Terremark's pricing starts at less than 4 cents an hour, for example, for a system with one virtual CPU and a half GB of memory, according to its Web site. Amazon's pricing starts at 10 cents per CPU hour.
Terremark's service is listed as being in beta. It wasn't immediately clear when the other companies' services would be rolled out.
VCloud Express supports VMware's goal of enabling its customers to move workloads from their own data centers to those of service providers — into the so-called public cloud — when they need additional computing capacity, and to move them back into their own data centers when they want to.