paboyle
8a79e93cc2
Rename the 5d domain wall fermion vectorised Ls impl class
2016-07-14 23:53:00 +01:00
paboyle
adbc7c1188
Adding files for multiple implementations (cache opt) and Ls vectorisation
...
of the 5D cayley form chiral fermions for the 5d matrix. With Ls entirely
in the vector direction, s-hopping terms involve rotations.
The serial dependence of the LDU inversion for Mobius and 4d even odd
checkerboarding is removed by simply applying Ls^2 operations (vectorised
many ways) as a dense matrix operation.
This should give similar throughput but high flops (non-compulsory flops)
but enable use of the KNL cache friendly kernels throughout the code.
Ls is still constrained to be a multiple of Nsimd, which is as much as 8 for AVX512
with single precision.
2016-07-14 22:59:21 +01:00
paboyle
ef97e32152
Adding persistent communicators
2016-07-08 17:16:08 +01:00
paboyle
a0676beeb1
Open up dependency on Eigen and FFTW
2016-07-07 22:31:07 +01:00
paboyle
680645f849
Merge branch 'release/v0.5.0'
2016-06-30 15:15:03 -07:00
paboyle
712b9a3489
Asm only for avx512
2016-06-30 14:35:02 -07:00
paboyle
bdaa5b1767
Updated to have perfect prefetching for the s-vectorised kernel with any cache blocking.
2016-06-30 14:35:02 -07:00
paboyle
8fcefc021a
Improved the prefetching when using cache blocking codes
2016-06-30 14:35:02 -07:00
paboyle
05c884a62a
Prefetch change
2016-06-30 14:35:01 -07:00
paboyle
2d8bb4c594
Tweaks
2016-06-30 14:35:01 -07:00
paboyle
6d58cb2a68
Enable reordering of the loops in the assembler for cache friendly.
...
This gets in the way of L2 prefetching however. Do next next link in stencil
prefetching.
2016-06-30 14:35:01 -07:00
Guido Cossu
5e02392f9c
Fixed compilation error for benchmark_dwf
...
Some parts were assuming floating point precision
2016-06-20 12:30:51 +01:00
paboyle
87418e7df1
Slightly faster prefetching perf.
2016-06-13 02:32:52 -07:00
paboyle
55f65b81b5
Improvements to the assembler interface that let us move chunks of the
...
site and s loop into the kernels. This will save on function call overhead and
guarantee L2 prefetching strategy is right since OMP can't distribute the
sub-chunks of work.
2016-06-09 01:12:36 -07:00
Azusa Yamaguchi
d9408893b3
Prefetching in the normal kernel implementation.
2016-06-08 05:43:48 -07:00
paboyle
8ac021de73
Added a test an fixed it for red black precon Ls innermost vectorised DWF
2016-06-07 13:16:56 -07:00
paboyle
e503ef5590
Cleaned up
2016-06-07 00:11:36 +01:00
paboyle
a7682b0060
Only instantiate the one routine to avoid duplicate symbol under g++5/MacOS
2016-06-06 23:48:21 +01:00
paboyle
53d06046b0
Compiling updates for KNL
2016-06-03 03:47:54 -07:00
paboyle
139cc5f1ae
Large change with KNL preparation
2016-06-03 03:24:26 -07:00
paboyle
5341977948
IMCI fixes. Thought I had committed these. The "real" disambiguation
...
between std::real and Grid::real shouldn't have been necessary and I don't
know why only the icpc v16.0 on babbage hits it.
May need a longer term rename of Grid::real or some careful EnableIf work.
2016-04-30 03:34:16 -07:00
paboyle
1e554350ac
The threaded coms didn't agree with GCC. Suprised, and looks like GCC bug.
2016-04-29 16:49:18 -07:00
paboyle
c79ea0dcef
Fixingn IMCI
2016-04-22 21:52:54 -07:00
paboyle
ba427abde9
simd 5d
2016-04-19 15:38:39 -07:00
paboyle
9b6ab6db16
simd in 5th dimension support
2016-04-19 15:38:01 -07:00
paboyle
806a83d38b
simd in fifth dim support for dwf
2016-04-19 15:36:19 -07:00
paboyle
b1192a8908
Benchmark_zmm added
2016-04-06 03:00:07 -07:00
paboyle
e8dddb1596
Adding extra benchmark
2016-04-06 10:32:54 +01:00
paboyle
e67fc2be18
Adding a trial for openmp overhead minimisation
2016-03-31 16:00:37 +01:00
paboyle
8052556275
Cleaning up the single/double kernel implementation switch
2016-03-31 14:51:32 +01:00
paboyle
60d965f79e
AVX512 improvements; sigfpe trapping too
2016-03-30 08:42:34 +01:00
paboyle
1ecbf9794d
Merge branch 'master' of https://github.com/paboyle/Grid
2016-03-30 08:37:55 +01:00
paboyle
c77b7ee897
AddSub based alternate SU3 routine
2016-03-28 17:55:22 -06:00
paboyle
1e355a51e1
Interface change
2016-03-27 23:46:55 -07:00
paboyle
21abaf7e91
Gamma sign change
2016-03-28 00:35:45 -06:00
paboyle
165bffc2e7
Avx512 changes for assembler kernels
2016-03-26 22:25:45 -06:00
paboyle
644fd6d32e
Build avx512 clean
2016-03-25 09:35:33 -07:00
paboyle
090e7aa930
Merge remote-tracking branch 'origin/chulwoo-dec12-2015'
...
Merge Chulwoo's Lanczos related improvements.
Merge Nd!=4 fixes for pure gauge HMC from Evan.
2016-03-08 09:55:14 +00:00
paboyle
325e745daa
Merge branch 'master' of https://github.com/paboyle/Grid
2016-03-02 07:04:03 -08:00
paboyle
61413565d0
Back off the inlined spin proj as not working
2016-03-02 07:03:09 -08:00
Antonin Portelli
497e7e4c53
BG/Q compatibility fix
2016-02-23 15:57:38 +00:00
Peter Boyle
6aeaf6f568
Parallel IO worked on. I'm puzzled because I already thought I shook this out on MacOS + OpenMPI and then
...
turned up problems on the BlueWaters Cray.
Gets 75MB/s from home filesystem on parallel configuration read. Need to make the RNG IO parallel,
and also to look at aggregating bigger writes for the parallel write.
Not sure what the home filesystem is.
2016-02-21 08:03:21 -06:00
Jung
9f0d9ade68
Added configure flag for LAPACK. Tested ImplicitlyRestartedLanczos::calc()
...
Checking in before cleaning up
2016-02-20 02:50:32 -05:00
paboyle
3425751cb8
Missing return value
2016-02-19 01:06:03 +00:00
Peter Boyle
22422a84d9
Small problem in compressor fix
2016-02-17 19:03:09 -06:00
Peter Boyle
c9fadf97a5
Simplify the compressor interface again.
2016-02-17 18:16:45 -06:00
Peter Boyle
81395e85d1
Regressing to not overlap comms and compute becasue bluewaters, edison, and cori are so rubbish at it.
2016-02-16 13:56:44 -06:00
Peter Boyle
a0fc47c6f9
Cheaper implementation
2016-02-15 16:02:36 -06:00
paboyle
e2f73e3ead
Updates for shmem
2016-02-10 16:50:32 -08:00
neo
6371676a75
Correcting some compilation errors for clang-sse
2016-02-10 11:37:03 +09:00
