Correlated Data Streaming
Tuesday April 17, 2007
Hamerschlag Hall D-210
Carnegie Mellon University
Research advocates address-correlating predictors to identify cache block addresses
for prefetch. Unfortunately, address-correlating predictors require correlation data
storage proportional in size to a program's active memory footprint. As a result,
current proposals for this class of predictor are either limited in coverage due to
constrained on-chip storage requirements or limited in prediction lookahead due to
long off-chip correlation data lookup.
In this talk, I will describe Last-Touch Correlated Data Streaming (LT-cords), a
practical address-correlating predictor. The key idea of LT-cords is to record
correlation data off chip in the order they will be used and stream them into a
practically-sized on-chip table shortly before they are needed, thereby obviating the
need for scalable on-chip tables and enabling low-latency lookup.
Mike is a third year PhD candidate in the Computer Architecture Laboratory at
Carnegie Mellon, where he is advised by Prof. Babak Falsafi. Mike's research
interests include processors and the software that runs on them. His work has
primarily been devoted to findings ways to improve processor performance through
hiding the latency of modern memory systems.