Peter Boyle
bfb68e6f02
Merge pull request #130 from giltirn/gparity-handunroll
...
Gparity handunroll
2017-09-21 10:11:00 +01:00
Christopher Kelly
59bd1fe21b
Fix for 'perm' and 'local' not being set for hand-unrolled external-site Dslash, which caused incorrect behavior of G-parity kernel
2017-08-29 13:07:37 -07:00
Christopher Kelly
74af885d4e
Removed some no-longer-needed associated with G-parity hand unrolled kernel
2017-08-29 09:50:37 -04:00
paboyle
80c5bce5bb
Merge branch 'develop' into feature/multi-communicator
2017-08-25 20:21:26 +01:00
paboyle
f68b5de9c8
No compile fix on Clang
2017-08-25 19:35:21 +01:00
Christopher Kelly
f365a83fae
In G-parity unrolled kernel, replaced calls to permute and exchange with run-time-evaluated permute type with explicit calls to appropriate underlying functions
2017-08-25 14:24:11 -04:00
Peter Boyle
c289699d9a
updated from cambridge mpi3 shakeout
2017-08-25 11:41:01 +01:00
Peter Boyle
c3b1263e75
Benchmark prep
2017-08-25 09:25:54 +01:00
Christopher Kelly
34a9aeb331
Reduced number of if-statement evaluations in G-parity unrolled kernel
2017-08-24 13:53:50 -07:00
Christopher Kelly
ce5df177ee
Removed superfluous implementation of G-parity twist for hand-unrolled kernel from GparityWilsonImpl
2017-08-23 15:05:22 -04:00
Christopher Kelly
a0bb8e5b46
Added hand-unrolled kernel implementations of all the other dslash precision / comms precision combinations with G-parity
2017-08-23 14:44:40 -04:00
Christopher Kelly
46f88e6d72
G-parity hand-unrolled intrinsics twist now uses one less permute and one less temporary
2017-08-23 13:21:10 -04:00
Christopher Kelly
b61835c1a5
Added inplace version of intrinsic G-parity twist to hand-unrolled kernel
2017-08-23 12:33:48 -04:00
Azusa Yamaguchi
d9cd4f0273
Staggered multinode block cg debugged. Missing global sum.
...
Code stalls and resumes on KNL at cambridge. Curious.
CG iterations 23ms each, then 3200 ms pauses. Mean bandwidth reports
as 200MB/s. Comms dominant in the report. However, the time behaviour suggests it
is *bursty*.... Could be swap to disk?
2017-08-23 15:07:18 +01:00
Christopher Kelly
061e48fd73
Replaced slow unpack-repack in G-parity BC twist with intrinsics version
2017-08-22 18:12:12 -04:00
Christopher Kelly
ab50145001
Implemented first, unoptimized version of hand-unrolled G-parity kernels
...
Improved Test_gparity
2017-08-22 17:12:25 -04:00
paboyle
a446d95c33
Trying to pass TeamCity and Travis
2017-08-20 01:10:50 +01:00
Peter Boyle
14d53e1c9e
Threaded MPI calls patches
2017-07-29 13:08:10 -04:00
paboyle
54e94360ad
Experimental: Multiple communicators to see if we can avoid thread locks in --enable-comms=mpit
2017-06-24 23:10:24 +01:00
7587df831a
Merge branch 'develop' into feature/hadrons
...
# Conflicts:
# lib/qcd/action/scalar/ScalarImpl.h
2017-06-20 15:50:39 +01:00
paboyle
46879e1658
Complex defined in Impl even for gauge.
2017-06-18 00:11:45 +01:00
0503c028be
Merge branch 'feature/qed-fvol' into feature/hadrons (non-trivial conflicts on scalar Impl)
...
# Conflicts:
# configure.ac
# lib/qcd/action/scalar/Scalar.h
2017-06-05 16:37:47 -05:00
Guido Cossu
9c12c37aaf
Confirming the fix on the complex boundary conditions
2017-05-09 08:41:29 +01:00
paboyle
529e78d43f
Restart the v0.7.0 release
2017-05-08 18:20:04 +01:00
paboyle
2439999ec8
Warning elimination; drop to -O2 on G++ bad versions
2017-05-06 14:44:49 +01:00
paboyle
1d96f662e3
Fixed 4d fermion gparity force. Put strong tests on make check force tests
2017-05-06 00:46:31 +01:00
Guido Cossu
20999c1370
Merge branch 'develop' into feature/hmc_generalise
2017-05-05 12:47:17 +01:00
paboyle
78ef10e60f
Mobius force improvement
2017-05-04 19:53:21 +01:00
paboyle
90f6bc16bb
No compile clang fix
2017-05-04 12:15:06 +01:00
Peter Boyle
422cdf4979
Some checks
2017-05-03 18:37:38 -04:00
Peter Boyle
38db174f3b
Print statement
2017-05-03 18:25:26 -04:00
Guido Cossu
4063238943
Adding HMC test file example for Mobius + smearing
2017-05-01 13:44:00 +01:00
Guido Cossu
3344788fa1
Merge branch 'develop' into feature/hmc_generalise
2017-05-01 12:13:56 +01:00
Peter Boyle
99220f6531
Fixes and better timing
2017-04-26 17:24:11 -04:00
Peter Boyle
f8797e1e3e
bug fix. works now and great face performance
2017-04-26 03:14:02 -04:00
Peter Boyle
fd1eb7de13
Clean implementation of the exterior faces listing only those points on the boudary
2017-04-26 02:34:52 -04:00
Peter Boyle
2ce898efa3
Pretty code
2017-04-26 02:34:25 -04:00
paboyle
ab66bac4e6
Think I'm getting on top of the reduced cost exterior precomputed list of links
2017-04-25 08:50:26 +01:00
paboyle
56277a11c8
Build a list of whats on the surface
2017-04-24 17:06:15 +01:00
Peter Boyle
5b55867a7a
Slightly cheaper Ext assembly
2017-04-24 05:36:11 -04:00
Peter Boyle
3accb1ef89
Debugged assemply split phase with interior suppression
2017-04-23 19:30:19 -04:00
Peter Boyle
e3d0e31525
Debugged assemply split phase with interior suppression
2017-04-23 19:29:27 -04:00
Peter Boyle
5812eb8a8c
Partially fixed. But the comms-overlap does not work yet.
2017-04-22 18:50:25 -04:00
paboyle
ac58565d0a
Dangerous rewrite of the assembly. If I make a mistake the debug will be painful.
2017-04-22 19:31:04 +01:00
paboyle
b722889234
Try a better load balancing loop
2017-04-22 19:27:41 +01:00
paboyle
abba44a837
Hand unrolled for overlapped comms
2017-04-22 17:45:17 +01:00
paboyle
f301be94ce
Fixed
2017-04-22 17:42:31 +01:00
Peter Boyle
1d1b225497
Hand unrolled Nc=3 kernels support split phase compute (on-node, off-node).
2017-04-22 09:05:28 -04:00
Peter Boyle
53a785a3dd
Fixing the KNL compile
2017-04-22 08:11:51 -04:00
paboyle
736bf3c866
Major rework of stencil. Half precision and MPI3 now working.
2017-04-22 11:33:50 +01:00
