Abstract

In a modern world for non numeric applications, out-of-order super scalar processors are designed to achieve higher performance. Unfortunately the improvement in the performance has lead to the increase in the chip power and energy dissipation. The load/store queue is a one of the major power consuming unit in the data path design during dynamic scheduling. Load/store queue is designed to absorb busts in cache access and maintain the order of memory operations by keeping all in-flight memory instruction in program order. The proposed technique aims at reducing both dynamic and static power dissipation in the load/store queue (LQ/SQ) by using power-gating technique and priority encoder. Through this implementation, the least amount of redesign, verification efforts, lowest possible design risk, least hardware overhead is achieved without significant impact on the performance.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call