R. F. DeMara and H. Kitano,
"PACE Benchmark Set," in Proceedings of the 1991 International Joint Conference 
on Artificial Intelligence (IJCAI'91) - Workshop on Parallel Processing for AI, 
pp. 517 - 520, Sydney, Australia, August 24 - 25, 1991.

Abstract:
We address architectural evaluation of massively parallel machines suitable for 
artificial intelligence. Our approach was to identify the impact of the specific 
algorithms features by measuring execution time on SNAP-1 and Connection 
Machine-2 using different knowledge base and machine configuration. Since a wide 
variety of parallel AI languages and processing architectures are currently in 
use, we developed a portable benchmark set for Parallel AI Computation 
Efficiency. PACE provides a representative set of processing workloads, 
knowledge base topologies, and performance indices. We also analyze speedup and 
scalability of fundamental AI operations in terms of the massively parallel 
paradigm.