GENERAL INFORMATION:
The CHOLESKY program performs blocked Cholesky Factorization on a sparse
matrix. The implementation contained in SPLASH-2 is described in:
Rothberg, E. Exploiting the Memory Hierarchy in Sequential and Parallel
Sparse Cholesky Factorization. Ph.D. Thesis, Stanford University,
November 1992.
This program works under both the Unix FORK and SPROC models.
RUNNING THE PROGRAM:
To see how to run the program, please see the comment at the top of the
file solve.C, or run the application with the "-h" command line option.
A set of input data files is included in the subdirectory "inputs." The
size of the cache (in bytes) should be specified on the command line, as
well as the number of processors being used. The postpass partition size
should be kept at the default value of 32. If this parameter is changed,
the value should be reported in any results that are presented.
BASE PROBLEM SIZE:
The base problem size for an upto-64 processor machine is the input file
tk29.O. The cache size (in bytes) should be specified on the command
line, as well as the number of processors being used. The postpass
partition size should not be changed. If it is, however, the new value
should be reported in any results that are presented.
DATA DISTRIBUTION:
Our "POSSIBLE ENHANCEMENT" comments in the source code tell where one
might want to distribute data and how. Data distribution has an impact
on performance on the Stanford DASH multiprocessor.
Sanchayan Maity
0f4b39775c