Title : Performance Studies of a Speculative Prefetch Model for
                Distributed Information Systems

Prof. Mohan Kumar
School of Computing Science
Curtin University of Technology
Perth Western Australia


Abstract: We investigate speculative prefetching under a model in
 which prefetching gets equal priority in resource
 utilisation. Theoretical limit for improvement in access time due to
 prefetching under uniform retrieval time.  For nonuniform retrieval
 time we evaluate mainline and branch prefetch. The improvement in
 access time is expressed as a stretch knapsack problem and an
 algorithm to maximise the improvement in accesstime is developed.
 Integration between speculative prefetching and caching is under
 investigation.



Brief biosketch of M Kumar

Currently working as a senior lecturer at Curtin University.  Prior to
joining Curtin University in 1992, worked as a Scientific Officer at
the Indian Institute of Science.  Obtained the Master of Technology
and PhD degrees from the Indian Institute of Science in 1985 and 1992
respectively.  Was the recepient of the gold medal for the PhD thesis
for the year 1992-93. Has published over 50 articles in refereed
journals and conference proceedings in the areas of Parallel and
Distributed Computing and Mobile Computing.

Note: The work was done in collaboration with N J Tuah and S Venkatesh