ec24a1f828
Fixed 2 bugs in LapEvec: 1) InsertLocalSlice 2) ensure convergence assertion stops entire machine
2019-05-03 16:03:56 +01:00
0efe63f6fa
3D smearing fix
2019-05-02 19:37:59 +01:00
b7ead6c16a
Fixed bug: iff stout smearing disabled then gauge field uninitialised
2019-05-02 18:20:49 +01:00
d9438627d9
M5D benchmark without vector copy overhead
2019-05-02 11:10:57 +01:00
b23305dbe2
fix M5D flop count
2019-05-02 11:08:21 +01:00
d3b5c02e2d
measure M5D bandwidth and fix M5D flop count
2019-05-02 11:02:39 +01:00
8b6541fb60
Fix gpu MultRealPart and MaddRealPart bug
2019-05-02 10:58:17 +01:00
6da9aa9971
replace std::vector with Vector in benchmark
2019-05-02 10:56:22 +01:00
44e0360b97
replace std::vector with Vector
2019-05-02 10:55:36 +01:00
9003c4a07c
allocator copy constructor (to be fixed)
2019-05-02 10:53:37 +01:00
b52fa38f8c
seed initialisation of RNG5
2019-05-02 10:36:09 +01:00
3f1c4d8789
fix comment hash
2019-05-02 10:24:36 +01:00
62692b68b9
I'd forgotten that Intel '17 doesn't like auto var{value}; syntax
2019-05-01 20:45:16 +01:00
311c35a15c
Looking for fixes for Intel '17 compiler errors. std::cout << complex number ?
2019-05-01 18:22:08 +01:00
a3fe57f430
NamedTensor writes to tag NamedTensor by default (not filename) - so still usable in case user renames file.
...
Also tweaked tensor index name checking (which is used to ensure tensor is correct type)
2019-05-01 18:11:37 +01:00
8dc0587621
Post Michael / Felix review. Ready for Peter / Antonin review
2019-05-01 13:04:51 +01:00
cfe5fa7a35
1) Don't write Laplacian eigenvectors to disk 2) Add a test that loads perambulators from disk
2019-05-01 09:50:23 +01:00
e72e26c899
Get rid of unnecessary multiFile options
2019-05-01 08:53:08 +01:00
334f29becb
Fairly close to ready for release. Felix and I to review, then submit for release
2019-04-30 23:53:57 +01:00
e56ead55ef
WIP
2019-04-30 14:41:48 +01:00
guelpers
4f0631615f
A2A Lepton-Meson Field contraction
2019-04-30 12:04:59 +01:00
d74d443d1b
Pre-release cleanup in progress
2019-04-29 22:18:29 +01:00
4203105104
Part-way through release tidy-up
2019-04-29 18:40:38 +01:00
guelpers
c2cd0e15d7
Merge remote-tracking branch 'upstream/develop' into feature/kl2QED
...
Conflicts:
Grid/qcd/action/fermion/DomainWallFermion.h
Grid/qcd/action/fermion/FermionOperator.h
2019-04-29 12:07:20 +01:00
ac19c0e04f
This will need to be removed eventually, but should save us fiddling about with each release
2019-04-29 09:20:08 +01:00
b48ca8a6ef
Merge branch 'develop' into feature/distil
...
* develop: (36 commits)
Mobius 2+1f sign off.
Integrator logging on by default
RHMC for mobius
HMC make file
Update
Simple check
Simple checks
Monius HMC
Changes locally
Power method
Momentum rescaling
Bounds checking
Bounds checking
Scale momentum convention to CPS/UKQCD MD time
Add bounds checking
Updated documentation after Peter's review. 1) Removed version numbers from Grid dependencies 2) Explained in a little more detail how to use Xcode to build Grid and Hadrons libraries
Remove bundled Eigen stuff
Fix typo so it matches develop
Remove bundled source from my local repository
Slightly generalize interface to SchurRedBlackBase and derived solver classes so we can pass forecasted initial guesses in EOFA heatbath correctly
...
2019-04-29 08:37:39 +01:00
c48ae4f3ad
1) Only the boss should write the perambulator - possibly was a source of intermittent corruption?
...
2) Implemented and test a perambulator conversion utility in Test_distil (commented out near the start of main)
2019-04-28 23:24:57 +01:00
fb74de0798
Making sure Hdf5 is an optional dependency (default to binary writer if not present)
2019-04-28 20:23:44 +01:00
adc1eaee68
Switched to Hdf5 format for perambulators. Ready for first test on Tesseract.
2019-04-28 17:53:42 +01:00
60330e05a3
NVCC wacky compiler options frozen. Possibly Cuda 9.2 specific
2019-04-28 07:39:33 +01:00
f9b8c0cccf
Vector changes for UVM
2019-04-28 07:38:57 +01:00
3cad67e569
Compile on tesseract
2019-04-28 07:38:09 +01:00
170ba4e619
Ensure different MPI ranks use different GPUs. The mapping works on Tesseract.
2019-04-28 07:32:30 +01:00
204a090497
Inner product is not working on GPU. Why?
2019-04-28 07:31:56 +01:00
3c717c47ef
GPU no compile on Wilson Multigrid fixed
2019-04-28 07:31:19 +01:00
5aca4e8670
Just realised that the trace is at every lattice site, so moved the check for no smearing further up
2019-04-26 17:23:18 +01:00
e223d0b99f
Need to validate range about which exp^iQ is considered unity
2019-04-26 16:00:35 +01:00
2e220456d3
First attempt at minimising smearing
2019-04-26 15:54:05 +01:00
4333d97958
fixed parameter
2019-04-26 14:29:21 +01:00
55c9c45d4b
Merge branch 'feature/distil' of https://github.com/mmphys/Grid into feature/distil
2019-04-26 14:28:01 +01:00
e70e03f560
started stout smearing for small w
2019-04-26 14:27:40 +01:00
ff5e2e0f47
Debug output fix. Meant to print the rho matrix for stout smearing ... not the address of the function that creates it
2019-04-26 12:30:41 +01:00
4f3d1ea6e8
Two heads are better than one. Combined effort and hopefully spatial smearing now fixed!
2019-04-26 12:18:11 +01:00
b1768ba820
Urgh!
2019-04-26 10:04:27 +01:00
3ac5a69a57
Ready to test spatial smearing (again)
2019-04-26 08:54:30 +01:00
50a74eaea3
Doesn't compile. Does it still need to be maintained?
2019-04-26 08:33:10 +01:00
8419fbb335
Renamed PerambLight module. Check with Felix whether Test_24 and Test_tesseract still need to be maintained
2019-04-26 08:23:15 +01:00
23a9b93cda
More dependencies for Distil.hpp move and (C) 2019 only
2019-04-26 07:39:05 +01:00
ecdc3ddebf
Moved Distil.hpp and added GNU license to all files
2019-04-26 07:24:56 +01:00
df41de4cb6
Merge branch 'develop' of https://github.com/paboyle/Grid into develop
2019-04-24 12:02:50 +01:00
Michael Marshall
gfilaci
guelpers
Peter Boyle
Felix Erben
fionnoh