Slow memory: the rising cost of optimism
R.A. Meyer, Jay Martin, Rajive L. Bagrodia · 2000
Rapid progress in the design of fast CPU chips has outstripped progress in memory and cache performance. Optimistic algorithms would seem to be more vulnerable to poor memory performance because they require extra memory for state saving and anti-messages. We examine the performance of both optimistic and conservative protocols in controlled experiments to evaluate the effects of memory speed and cache size, using a variety of applications.