Using analytics to optimize cloud computing performance and cost savings

IBM inventors have patented a technique that uses analytics to increase cloud computing performance and reduce costs by dynamically moving workloads—between or within cloud centers—based on an automatic analysis that determines the most efficient and effective use of available resources.

IBM's patented invention automatically pools cloud computing resources from disparate cloud centers, enabling clients to improve performance and save money by reducing the amount of manual intervention needed to allocate hardware, software and services. The automated technique enables users have more control over where and how their cloud workloads are run and facilitates optimal use of a cloud infrastructure.

IBM received U.S. Patent #8,676,981 B2 "Routing Service Requests Based on Lowest Actual Cost within a Federated Virtual Service Cloud" for the invention.

"Cloud delivery models are evolving and maturing and are creating new possibilities for end-users and third-party service providers to lower costs for IT deployments as they enter new markets and create new solutions," said Leonard Hand, IBM Master Inventor and lead inventor on the patent. "Our patented invention will give users more control over their cloud deployments so that they can extract the most value from their IT resources at the most optimal cost."

Major cloud infrastructure providers like IBM are building global networks of cloud centers to deliver local services to clients around the world. Patent #8,676,981 can help streamline the management of local and geographically dispersed cloud hubs, while improving application performance and responsiveness by lowering latency of cloud deployments.

IBM's patented dynamic cloud management invention could also be used by Managed Service Providers (MSPs) and channel resellers who leverage cloud infrastructure from third-party providers to build and deliver value-added solutions. MSPs could use IBM's invention to isolate and automatically support their customers' specific workload demands and truly enable usage-based-consumption pricing models.