Big data sets require big compute power. Machine translation, speech recognition and video processing are some examples of these big data sets. Since they are so large, these big data sets usually require parallel processing and parallel storage systems, often using clustered resources that are accessed over the web. This is the basic model that is referred to as Cloud Computing. Researchers at Intel’s lab in Pittsburgh have two initiatives that support this cloud computing paradigm – OpenCirrus and Tashi.

OpenCirrus is a joint initiative started by Intel, HP and Yahoo as well as academic partners. In a back room of the lab, server racks with nearly 1000 cores were running the Tashi operating environment software that manages the 400 terabytes of resources in the cluster. Individuals and groups can submit proposals to utilize these large clusters to run experiments and develop applications.