portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-06-16 14:57:05 +01:00
Code Releases Activity

Sizable improvement in multigrid for unsquared.

Browse Source 
6000 matmuls CG unprec
2000 matmuls CG prec (4000 eo muls)
1050 matmuls PGCR on 16^3 x 32 x 8 m=.01

Substantial effort on timing and logging infrastructure
This commit is contained in:
Peter Boyle
2015-07-24 01:31:13 +09:00
parent 11c99d5e66
commit d1afebf71e
67 changed files with 945 additions and 753 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
lib/lattice/Lattice_transfer.h
View File

@ -23,7 +23,7 @@ inline void subdivides(GridBase *coarse,GridBase *fine)
template<class vobj> inline void pickCheckerboard(int cb,Lattice<vobj> &half,const Lattice<vobj> &full){
half.checkerboard = cb;
int ssh=0;
PARALLEL_FOR_LOOP
//PARALLEL_FOR_LOOP
for(int ss=0;ss<full._grid->oSites();ss++){
std::vector<int> coor;
int cbos;
@ -40,7 +40,7 @@ PARALLEL_FOR_LOOP
template<class vobj> inline void setCheckerboard(Lattice<vobj> &full,const Lattice<vobj> &half){
int cb = half.checkerboard;
int ssh=0;
PARALLEL_FOR_LOOP
//PARALLEL_FOR_LOOP
for(int ss=0;ss<full._grid->oSites();ss++){
std::vector<int> coor;
int cbos;
@ -158,6 +158,7 @@ template<class vobj,class CComplex>
fine_inner = localInnerProduct(fineX,fineY);
blockSum(coarse_inner,fine_inner);
PARALLEL_FOR_LOOP
for(int ss=0;ss<coarse->oSites();ss++){
CoarseInner._odata[ss] = coarse_inner._odata[ss];
}