The file paper.ps contains the CSRI technical report: 

   Coordinated Allocation of Memory and Processors in Multiprocessors

                                  by

                 Eric W. Parsons and Keneth C. Sevcik

If you have the UNIX uncompress program, get the paper.ps.Z file.
Remember to transfer the file in binary mode.  Uncompress it, and 
print it on a postscript printer.

If you do not have uncompress, get the paper.ps file in ascii mode, 
and print it on a postscript printer.

This report was made available for anonymous ftp by: 

                         eparsons@cs.toronto.edu

ABSTRACT 

  An important issue in multiprogrammed multiprocessor systems is the
  scheduling of parallel jobs.  Most research in the area has focussed
  solely on the allocation of processors to jobs.  However, since memory is
  also a critical resource for many parallel jobs, the allocation of memory
  and processors must be coordinated to allow the system to operate most
  effectively.

  To understand how to design such coordinated scheduling disciplines, it
  is important to have a theoretical foundation.  To this end, we develop
  bounds on the achievable system throughput when both memory and
  processing time are in demand.  We then propose and simulate a simple
  discipline and relate its performance to the throughput bounds.  An
  important result of our work is for the situation in which the workload
  speedup is convex (from above), but the speedup characteristics of
  individual jobs are unknown.  It shows that an equi-allocation strategy
  for processors can achieve near-maximum throughput, yet offer good mean
  response times, when both memory and processors are considered.