Microsoft officials are seeing more researchers, who are trying to push the boundaries of scientific discovery but often come to the realization that there are too few supercomputers to go around, turning to cloud computing environments as a solution.
Dennis Gannon, director of cloud research strategy for Microsoft’s Research division, said that clouds are providing a vital, cost-effective alternative to grids and supercomputers.
Gannon reported in a company blog post that following an informal survey of scientific research projects conducted on Azure, Microsoft’s cloud computing platform, the software company discovered that 90 percent of those polled “were pleased with their ROI using cloud services to build their application and would use cloud resources again.”
He cautioned—and business users should note—that the sample is biased since as a rule, these researchers can be considered “leading-edge risk takers and early adopters.” Nonetheless, Azure is helping to put large-scale computing resources within reach of scientists that are largely locked out of the supercomputing realm or the high-performance computing (HPC) scene.
As evidence, Gannon cited a research program from the Geography of Natural Disasters Laboratory at the University of Aegean in Greece. Researchers built a cloud application that can simulate wildfire propagation, giving emergency-response agencies on Lesvos a potentially life-saving tool for dealing with the island’s wildfires.
Azure first debuted in 2008 to compete with Amazon Elastic Cloud Compute (EC2), Google App Engine and other cloud computing upstarts. Meant to help usher in the company’s “software plus services” era, Azure—formerly Project Red Dog—was touted as a developer-friendly shortcut to the burgeoning business cloud market. Microsoft boasted that Windows developers could quickly float their own cloud apps, courtesy of Visual Studio, the company’s popular Windows development environment.
Today, Azure serves as the lynchpin to many of the company’s business cloud services. In the past year, Microsoft has been bulking up its cloud storage capabilities by updating SkyDrive, inking deals with Symantec for cloud-based disaster recovery and ultimately snapping up StorSimple for hardware appliances that bridge on-premises storage systems and Azure’s cloud storage.
During the Windows Server 2012 launch in September, company officials blurred the lines of on-premises software by dubbing it the company’s “Cloud OS.” For instance, one of the virtualization-friendly, cloud-enabled operating system’s many new features is the ability to sync in-house Active Directory credentials with Windows Azure Directory.
While Microsoft has been busy making an enterprise IT case for Azure during the past four years, as a Cloud Research Engagement Initiative supporter, it has also been letting the scientific community leverage the company’s massive cloud infrastructure. To date, the program’s partners have awarded over 75 research teams time on Azure.
Cloud services also stand uniquely positioned to deal with the challenges that advanced computing platforms pose to research institutions, said Gannon.
“The vast majority of scientists don’t want to learn to program a cluster, a supercomputer or a cloud. They want to get on with their science. This describes the vast majority of the research community,” wrote Gannon.
As for the future, Microsoft has opened a dialogue with the research community and is working with Internet2 to ultimately create self-sustaining, cloud-based scientific research ecosystem. “In the cloud world, we have infrastructure as a service, platform as a service and software as a Service. Why not research as a service?” Gannon said.
As for the future, Microsoft has opened a dialogue with the research community and is working with Internet2 to ultimately create self-sustaining, cloud-based scientific research ecosystem. “In the cloud world, we have ‘Infrastructure as a Service,’ ‘Platform as a Service’ and ‘Software as a Service.’ Why not ‘Research as a Service’?” Gannon wrote.