Software Based Approach : Hardware

1713 Words7 Pages
Software Based Approach: Hardware based approaches constitute only a part of the solution, they offer limited power knob options. Considering the continuously evolving hardware configurations and increasing hardware heterogeneity, data management software shall play an important role in energy optimization. Query optimization and physical data independence provide opportunities for software level control over power performance tradeoffs [7,32]. Adapting Query Optimizer Energy optimization has emerged as a new parameter for query optimization. Research in this area is targeted at improving the query optimizer component to compute and select query plans taking into consideration both SLA based performance goals and energy efficiency. To lower the energy consumption, query optimizers need accurate power models to estimate energy costs. Power cost is calculated by obtaining specifications of hardware components and dividing & computing related estimated time through an iterative approach. In [23], authors proposed ERP (Energy Response Time Profile), a structure that contains a detailed energy and response time cost of executing each query plan at every possible power/performance setting under consideration. The ERP of a query can then be used to select the appropriate “energy-enhanced” strategy to execute the query. This paper [23] mainly focuses on taking advantage of the slack (difference between optimal performance and SLA goal) and leveraging it to reduce
Open Document