R3-DLA (Reduce, Reuse, Recycle): A More Efficient Approach to Decoupled Look-Ahead Architectures

Sushant Kondguli, Michael Huang

Modern societies have developed insatiable demands for more computation capabilities. Exploiting implicit parallelism to provide automatic performance improvement remains a central…