Profit-Driven Service-Chain Deployment For EDA Requests On Private Cloud

Abstract

Cloud computing is becoming a pervasive technology in many fields. However, the EDA field has not yet developed a working prototype that employs the commercial tools as services and deploys them on the cloud. Therefore, in this paper, we aim at three core problems: request acceptance, service schedule, and resource deployment on private EDA clouds, and give an integrated solution for processing requests. Compared to Conventional Inclusion (C.I.), total profit can increase under a high arrival rate by 29% after applying Progressive Inclusion (P.I.). Meanwhile, the non-preemptive, penalty-considered, Least Slack First Scheduling deals with the order of services to be scheduled. As to resource deployment, parcel-fit (PF) outperforms best-fit (BF) and first-fit (FF) by 5% and 18% on total profits under a high arrival rate. To sum up, the proposed solution provides an effective way to build a private EDA cloud with more profits than the conventional solution.