portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2026-07-17 15:43:27 +01:00
Code Releases Activity
3,890 Commits 95 Branches 14 Tags
e3a309a73f37cadad41b031a13faae02a6f6b915
Commit Graph
100 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
paboyle e3a309a73f Eigen happiness 2018-07-31 00:35:17 +01:00
paboyle 65533741f7 7 moms 2018-07-28 16:17:47 +01:00
paboyle 44f4f5c8e2 Momentum loop 2018-07-27 23:00:16 +01:00
paboyle 71e1006ba8 Updated meson field benchmark for dirac structures 2018-07-26 09:09:29 +01:00
paboyle ec9939c1ba Test for faster implementation of meson field inner loop 
This should be possible to cache block at outer levels, global sum across nodes not performed
and deferred to caller to block them all into a big all reduce.
Nc=3 and Fermion is hard coded in an ugly way. We might think about benchmarking whether
a product without the conjugate should be made available by Grid.

It is not clear whether the explicit unroll, or the performing of conjugate on left once
was the real source of the speed up.

Gives 70-80 GF/s on my laptop (single) half that double, and 70GB/s to cache.

This is competitive with dslash and a reasonable stopping point for the optimisation. If necessary we can revisit.
 2018-07-10 12:38:51 +01:00
paboyle 5dfd216a34 Better thread safety 2018-06-04 21:08:44 +01:00
paboyle c2e8d0aa88 Solve g++ problem on the lanczos test 2018-06-04 18:34:15 +01:00
paboyle bf96a4bdbf Merge branch 'master' into develop 2018-06-04 14:03:11 +01:00
paboyle 84685c9bc3 Overflow fix 2018-06-04 13:42:07 +01:00
paboyle 6eed167f0c Merge branch 'release/0.8.1' 2018-05-04 17:34:11 +01:00
paboyle 4ad0df6fde Bump volume for Gerardo 2018-05-04 17:33:23 +01:00
Peter Boyle 68a5079f33 Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-05-04 14:13:54 +01:00
Peter Boyle 8634e19f1b Update 2018-05-04 14:13:35 +01:00
Peter Boyle 8c658de179 Compressor speed up (a little); streaming stores 2018-05-02 17:52:16 +01:00
Peter Boyle 6f6c5c549a Split off gparity 2018-05-02 14:11:23 +01:00
Peter Boyle 1584e17b54 Revert to fast versoin 2018-05-02 14:10:55 +01:00
Peter Boyle 12982a4455 Hypercube optimisation 2018-05-02 14:10:21 +01:00
Peter Boyle 172f412102 shmget reintroduce 2018-05-02 14:07:41 +01:00
Peter Boyle a64497265d TIming 2018-05-02 14:07:28 +01:00
Peter Boyle c45f24a1b5 Improvements for tesseract 2018-04-30 21:50:00 +01:00
paboyle 661f1d3e8e Merge branch 'release/0.8.0' into develop 2018-04-27 11:22:33 +01:00
paboyle edcf9b9293 Merge branch 'release/0.8.0' 2018-04-27 11:13:19 +01:00
paboyle fe6860b4dd Update with LIME library guard 2018-04-27 08:57:34 +01:00
paboyle d6406b13e1 Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-04-27 07:52:56 +01:00
paboyle e369d7306d Rename 2018-04-27 07:51:44 +01:00
paboyle 9f8d63e104 Roll over version 2018-04-27 07:51:12 +01:00
paboyle 9b0240d101 Hot start test 2018-04-27 07:50:51 +01:00
paboyle b27f0e5a53 Control over IO 2018-04-27 07:50:15 +01:00
paboyle 75e4483407 Stronger convergence test 2018-04-27 07:49:57 +01:00
paboyle 809b1cdd58 Bug fix for MPI running ; introduced last night 2018-04-27 05:19:10 +01:00
paboyle 1be8089604 Clean compile 2018-04-26 23:42:45 +01:00
paboyle 3e0eff6468 Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-04-26 23:00:46 +01:00
paboyle 7ecc47ac89 Quenched test compile 2018-04-26 23:00:28 +01:00
paboyle e9f1ac09de static 2018-04-26 23:00:08 +01:00
Peter Boyle fa0d8feff4 Performance of CovariantCshift now non-embarrassing. 2018-04-26 17:56:27 +01:00
Peter Boyle 05b44aef6b Merge branch 'develop' of https://github.com/paboyle/Grid into develop 
Conflicts:
	benchmarks/Benchmark_su3.cc
 2018-04-26 15:38:49 +01:00
Peter Boyle 03e9832efa Use macros for bare openmp 2018-04-26 14:50:02 +01:00
Peter Boyle 28a375d35d Force static 2018-04-26 14:49:42 +01:00
Peter Boyle 3b06381745 Guard bare openmp statemetn with ifdef 2018-04-26 14:48:57 +01:00
Peter Boyle 91a0a3f820 Improvement 2018-04-26 14:48:35 +01:00
Peter Boyle 8f44c799a6 Saving the benchmarking tests for Cshift 2018-04-26 14:48:03 +01:00
paboyle 2baf193031 Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-04-25 00:14:03 +01:00
paboyle 362ba0443a Cshift updates 2018-04-25 00:12:11 +01:00
paboyle 276a2353df Move constructor 2018-04-25 00:11:07 +01:00
paboyle 870b1a85ae Think I have the physical prop interface to CF and PF overlap right, but need a strong check/regression. 
Only support Hw overlap, not Ht for now. Ht needs a new Dminus implemented.
 2018-04-18 14:17:49 +01:00
paboyle b5510427f9 physical fermion interface, cshift benchmark in SU3. 2018-04-18 01:43:29 +01:00
paboyle f7f043d8cf Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-04-17 10:57:18 +01:00
paboyle ddcaa6ad29 Master does header on Nersc 2018-04-17 10:48:33 +01:00
paboyle f089bf5629 Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-03-30 16:17:26 +01:00
paboyle 276f113f28 IO uses master boss node for metadata. 2018-03-30 16:17:05 +01:00
paboyle ab6afd18ac Still compile if no LIME 2018-03-30 13:39:20 +01:00
paboyle 5bde64d48b Barrier required in parallel when we use ftell 2018-03-30 12:41:30 +01:00
paboyle 2f5add4d5f Creation of file 2018-03-30 12:30:58 +01:00
paboyle 5436308e4a Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-03-21 14:26:29 +00:00
paboyle 07fe7d0cbe Save file in current dir; print checksums 2018-03-21 14:26:04 +00:00
paboyle 954e38bebe Put a username in the path 2018-03-20 18:16:15 +00:00
paboyle b1a38bde7a Extra test for Gparity with plaquette action 2018-03-20 18:01:32 +00:00
Peter Boyle 6c6d43eb4e Drop RB on coarse space ; that was a mistake 2018-03-17 09:35:01 +00:00
Peter Boyle e1dcfd3553 typo fix 2018-03-16 23:10:47 +00:00
Peter Boyle 888838473a 4GB clean the offsets in parallel IO for multifile records 2018-03-16 21:54:56 +00:00
Peter Boyle 01568b0e62 Add a new SHM option 2018-03-16 21:54:28 +00:00
Peter Boyle d5ce66f6ab Extra SHM option 2018-03-16 21:37:03 +00:00
paboyle 0880747edb Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-03-09 20:44:42 +00:00
paboyle b801e1fcd6 fclose should be called through a call to close() 2018-03-09 20:44:10 +00:00
paboyle c399c2b44d Guido broke the charge conjugate plaquette action with premature optimisation. 
This sector of the code does not matter for anything other than Guido's quenched HMC
studies, and any plaq specific optimisations should be retained in a private branch
instead of destroying the code simplicity.
 2018-03-05 12:55:41 +00:00
paboyle af7de7a294 Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-03-05 12:22:41 +00:00
paboyle 1dc86efd26 Finalize protection 2018-03-05 12:22:18 +00:00
paboyle 4790e99817 Extra communicator free that I had missed. 
Hard to audit them all as this is complex
 2018-02-20 15:12:31 +00:00
paboyle 2dd63aa7a4 Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-02-20 14:29:26 +00:00
paboyle 559a501140 Deflation interface for solvers 2018-02-20 14:29:08 +00:00
paboyle 945684c470 updates for deflation in the RB solver 2018-02-20 14:28:38 +00:00
paboyle c96483e3bd Whitespace only change 2018-02-13 11:39:07 +00:00
paboyle ae31a6a760 Move deflate to right class 2018-02-13 02:11:37 +00:00
paboyle dd8f2a64fe INterface to suit hadrons on Lanczos 2018-02-13 02:08:49 +00:00
paboyle 7b8b2731e7 Conj error for complex coeffs 2018-02-12 16:06:31 +00:00
paboyle 237a8ec918 Communicator leak fixed (I think) 2018-02-12 13:27:20 +00:00
Peter Boyle 896f3a8002 Fix to MPI for Hokusai system 2018-02-01 18:51:51 +00:00
paboyle e564d11687 Allow resize of the shared memory buffers 2018-01-08 15:20:26 +00:00
paboyle 0b2162f375 Clean up 2018-01-08 14:06:53 +00:00
paboyle 5610570182 Synthetic test of lanczos 2018-01-08 11:36:39 +00:00
paboyle 44f65526e0 Simplify communicators 2018-01-08 11:35:43 +00:00
paboyle 43e48542ab Merge branch 'develop' of https://github.com/paboyle/Grid into develop 2018-01-08 11:34:45 +00:00
paboyle 0b85f1bfc8 Simplify the communicator proliferation: mpi and none. 2018-01-08 11:33:47 +00:00
paboyle 9947cfbf14 Simplify number of communicator cases 2018-01-08 11:33:01 +00:00
paboyle 357badce5e Simplify communicator case proliferation 2018-01-08 11:32:16 +00:00
paboyle 0091eec23a Simplify communicator cases 2018-01-08 11:31:32 +00:00
paboyle 9e9c2962df Simplify comms layer proliferation 2018-01-08 11:30:22 +00:00
paboyle bda97212a9 Simplify proliferation of comms layers 2018-01-08 11:29:20 +00:00
paboyle b91282ad46 Simplify comms layer proliferation 2018-01-08 11:28:52 +00:00
paboyle 0a68470f9a Simplify comms layers 2018-01-08 11:28:30 +00:00
paboyle 6ecf280723 Simplify comms layer proliferation 2018-01-08 11:28:04 +00:00
paboyle 7eeab7f995 Simplify comms layers 2018-01-08 11:27:43 +00:00
paboyle 9b32d51cd1 Simplify comms layer proliferatoin 2018-01-08 11:27:14 +00:00
paboyle 7b3ed160aa Rationalise MPI options 2018-01-08 11:26:48 +00:00
paboyle 1a0163f45c Updated to do list 2018-01-08 11:26:11 +00:00
paboyle 9579c9c327 Threading improvement 2017-12-05 14:12:22 +00:00
paboyle 3729c7a7a6 Clean up of test 2017-12-05 13:07:31 +00:00
paboyle c24d4c8d0e Improved parallel RNG init 2017-12-05 13:01:10 +00:00
paboyle a14038051f Improved AllToAll asserts 2017-12-05 11:43:25 +00:00
paboyle 3e560b9462 Faster RNG init 2017-12-05 11:42:05 +00:00