ReDHiP: Recalibrating Deep Hierarchy Prediction for Energy Efficiency

Download Now
Provided by: University of Calgary
Topic: Hardware
Format: PDF
Recent hardware trends point to increasingly deeper cache hierarchies. In such hierarchies, accesses that lookup and miss in every cache involve significant energy consumption and degraded performance. To mitigate these problems, in this paper the authors propose Recalibrating Deep Hierarchy Prediction (ReDHiP), an architectural mechanism that predicts Last-Level Cache (LLC) misses in advance. An LLC miss means that all cache levels need not be accessed at all. Their design for ReDHiP focuses on a simple, compact prediction table that can be efficiently recalibrated over time. They find that a simpler scheme, while sacrificing accuracy, can be more accurate per bit than more complex schemes through recalibration.
Download Now

Find By Topic