reed@b.CS.UIUC.EDU (Daniel A. Reed) (03/15/86)
The 1987 Sigmetrics Conference on Performance Measurement and Modeling of Computer Systems plans several sessions specifically devoted to APPLICATIONS of performance modeling to computer systems. The Sigmetrics conference has traditionally been devoted to solution algorithms for queueing network models - and their application to computer systems. Applications related papers have been solicited but this has not been the conference theme. However, "times they are a changin'." This year's conference will have sessions devoted to Database/File Systems I/O Systems Locally Distributed Systems Networks Parallel Processing Supercomputing Someone working in each area is actively seeking papers. In this case, I (and Elizabeth Williams at Los Alamos) are the parallel processing and supercomputing representatives. Now, before you say, "I would not know a queueing network solution algorithm if it bit me," don't despair. To quote from the charge to solicitors, a good paper is one that applies performance evaluation techniques in a methodologically sound manner and is a good paper from the point of view of a researcher in the application area (i.e., it is a good parallel processing or supercomputing paper). Original research is required, but original performance evaluation techniques, or novel usage of old techniques, are NOT required. In a phrase, we are seeking the type of performance papers that appear in the Int'l Conference on Parallel Processing - performance evaluation as a means, not an end. The deadline for submission of papers is September 23, 1986. The conference will be held in May of 1987 in Banff, Alberta, Canada. We are soliciting papers from individuals whose work might be appropriate for the conference. Because we must determine if there is sufficient interest for a Parallel Processing and/or Supercomputing session at the conference, I NEED A TENTATIVE COMMITMENT TO SUBMIT A PAPER BY EARLY APRIL. THIS COMMITMENT IS NOT BINDING, HOWEVER. Moreover, failure to commit now does not preclude later submission. All papers submitted will be refereed, hence I cannot guarantee acceptance. But as the program committee member in charge, I will see that they get reviewed by knowledgeable people. If you have questions, or comments, please feel free to contact me at (217) 333-3807 (work) reed@b.cs.uiuc.edu arpanet reed@uiuc.csnet CSnet Or Elizabeth Williams at ew@lanl.arpa Or, failing all else: Dan Reed Department of Computer Science University of Illinois 1304 W. Springfield Ave Urbana, IL 61801