perm filename MULTIP.NOT[W82,JMC] blob sn#640289 filedate 1982-02-07 generic text, type C, neo UTF8
COMMENT āŠ—   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	Feb 7
C00003 ENDMK
CāŠ—;
Feb 7
	The ideas in my MULTIP[1,JMC]@S1 are probably not adequate
for programming subroutines.  Consider a matrix multiplication
subroutine.  Which parts are done in parallel and how much
parallelism is allowed needs to depend on the size of the matrices
in a more flexible way than I have allowed.