64-BitBrainstorm_2026/AMSS-NCKU
8
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
50e2a845f8c3e936e334514a66051082dcc41505
AMSS-NCKU/AMSS_NCKU_source
History
CGH0S7 50e2a845f8 Replace MPI_Allreduce with owner-rank MPI_Bcast in Patch::Interp_Points 
The two MPI_Allreduce calls (data + weight) were the #1 hotspot at 38.5%
CPU time. Since all ranks traverse the same block list and agree on point
ownership, we replace the global reduction with targeted MPI_Bcast from
each owner rank. This also eliminates the weight array/Allreduce entirely,
removes redundant heap allocations (shellf, weight, DH, llb, uub), and
writes interpolation results directly into the output buffer.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-09 22:39:18 +08:00
..
ABE.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
adm_constraint.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
adm_ricci_gamma.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Ansorg.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Ansorg.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Ansorg.psid
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
array.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
array.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
BH_diagnostics.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
BH_diagnostics.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Block.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Block.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn2adm.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_class.C
Optimize MPI communication in RestrictProlong and surface_integral
2026-02-09 22:07:12 +08:00
bssn_class.h
Optimize MPI communication in RestrictProlong and surface_integral
2026-02-09 22:07:12 +08:00
bssn_constraint.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_gpu_class.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_gpu_class.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_gpu_rhs_ss.cu
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_gpu.cu
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_gpu.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_macro.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_macro.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_rhs_ss.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_rhs.f90
Disable NaN sanity check in bssn_rhs.f90 for production builds
2026-02-06 18:36:29 +08:00
bssn_rhs.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssn_step_gpu.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssnEM_class.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssnEM_class.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssnEScalar_class.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssnEScalar_class.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
bssnEScalar_rhs.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cctk_Config.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cctk_Constants.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cctk_Types.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cctk.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cgh.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cgh.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
checkpoint.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
checkpoint.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
config.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
coords.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
coords.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cpbc_util.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cpbc.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cpbc.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cpm_map.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
cpm_map.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
DataCT.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
derivatives.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
diff_new_sh.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
diff_new.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
diff_newwb.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
driver.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
empart.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
empart.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
enforce_algebra.f90
Compiler-level and hot-path optimizations for GW150914
2026-02-06 17:13:39 +08:00
enforce_algebra.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
error_exit.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
expansion_Jacobian.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
expansion.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
fadmquantites_bssn.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
fadmquantites_bssn.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
fd_grid.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
fd_grid.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
FFT.f90
Optimize numerical algorithms with Intel oneMKL
2026-01-16 10:58:11 +08:00
find_horizons.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
fmisc.f90
Optimize polint/polin2/polin3 interpolation for cache locality
2026-02-06 19:00:35 +08:00
fmisc.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
fourdcurvature.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
fuzzy.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
gaussj.C
Optimize numerical algorithms with Intel oneMKL
2026-01-16 10:58:11 +08:00
getnp4.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
getnp4.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
getnp4EScalar.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
getnp4old.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
getnpem2.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
getnpem2.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
gfns.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
ghost_zone.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
ghost_zone.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
gpu_mem.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
gpu_rhsSS_mem.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
gr.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
horizon_sequence.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
horizon_sequence.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
ilucg.f90
Optimize numerical algorithms with Intel oneMKL
2026-01-16 10:58:11 +08:00
ilucg.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_guess.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_maxwell.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_maxwell.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_null2.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_null2.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_null.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_null.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_puncture.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_puncture.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_scalar.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
initial_scalar.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
IntPnts0.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
IntPnts.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Jacobian.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Jacobian.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
kodiss_sh.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
kodiss.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
kodiss.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
linear_map.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
linear_map.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
lopsidediff.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
macrodef.fh
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
macrodef.h
Switch compiler toolchain from GCC to Intel oneAPI
2026-01-15 16:32:12 +08:00
makefile
Only enable OpenMP for TwoPunctures
2026-02-08 13:00:37 +08:00
makefile.inc
PGO supported
2026-02-09 10:59:26 +08:00
misc.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
misc.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
miscfp.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
monitor.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
monitor.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
MPatch.C
Replace MPI_Allreduce with owner-rank MPI_Bcast in Patch::Interp_Points
2026-02-09 22:39:18 +08:00
MPatch.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
myglobal.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
MyList.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Newton.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
norm.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullEvol2.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullEvol.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullEvol.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullNews2.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullNews.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullNews.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullShellPatch2_Evo.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullShellPatch2.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullShellPatch2.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullShellPatch.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
NullShellPatch.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Parallel_bam.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Parallel_bam.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Parallel.C
Optimize MPI Sync with merged transfers, caching, and async overlap
2026-02-09 21:03:37 +08:00
Parallel.h
Optimize MPI Sync with merged transfers, caching, and async overlap
2026-02-09 21:03:37 +08:00
parameters.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_edge.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_info.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_info.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_interp.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_interp.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_system_info.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_system.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch_system.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
patch.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
perf.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
perf.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
point_diff_new_sh.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
prolongrestrict_cell.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
prolongrestrict_vertex.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
prolongrestrict.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
prolongrestrict.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
ricci_gamma.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
ricci_gamma.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
round.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
rungekutta4_rout.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
rungekutta4_rout.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
scalar_class.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
scalar_class.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
scalar_rhs.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
scalar_rhs.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
scalarwaves.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Set_Rho_ADM.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
setup.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
shellfunctions.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
shellfunctions.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
ShellPatch.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
ShellPatch.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
sommerfeld_rout.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
sommerfeld_rout.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
stdc.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
surface_integral.C
Optimize MPI communication in RestrictProlong and surface_integral
2026-02-09 22:07:12 +08:00
surface_integral.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
testNull2.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
testNull.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
tgrid.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
tgrid.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
tool_f.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
tool.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
transpbh.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
TwoPunctureABE.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
TwoPunctures.C
Use OpenMP's parallel for with schedule(dynamic,1)
2026-02-07 19:48:24 +08:00
TwoPunctures.h
Performance optimization for the TwoPunctures module
2026-02-07 14:48:47 +08:00
util_String.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
util_Table.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
util.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
var.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
var.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
writefile_f.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
z4c_rhs_point.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Z4c_class.C
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Z4c_class.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Z4c_rhs_ss.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
Z4c_rhs.f90
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
zbesh.for
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00
zbesh.h
asc26 amss-ncku initialized
2026-01-13 15:01:15 +08:00