[EMAIL PROTECTED] wrote on 12/13/2007 06:50:28 PM: > This reminds me of a similar issue I had. What approaches do you > take for large dense matrix multiplication in MPI, when the matrices > are too large to fit into cluster memory? If I hack up something to > cache intermediate results to disk, the IO seems to drag everything > to a halt and I'm looking for a better solution. I'd like to use > some libraries like PETSc, but how would you work around memory > limitations like this (short of building a bigger cluster)? >
Dear Peter, There are many algorithms for Matrix operations that depend on the properties of the matrix and the operation. You can easily add writing to a tmpfs RAM disk filesystem to speed methods that involve reading and writing of temporary files. So what I do now is take those old Fortran codes that read and write files and keep the intermediate result files in ramdisk. ------ Sincerely, Tom Pierce
_______________________________________________ Beowulf mailing list, [email protected] To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
