e4deea4b94
Weird bug appears with Vector<Vector<>>.
...
"fix" with std::vector<Vector<>>
Lies in the face table code. But think there is some latent problem.
Possibly in my allocator since it is caching, but could simplify or eliminate the caching
option and retest. One to look at later.
2018-09-11 04:36:57 +01:00
7bf82f5b37
Offload the face handling to GPU
2018-09-10 11:28:42 +01:00
704ca162c1
Offloadable compression
2018-09-10 11:20:50 +01:00
da17a015c7
Pack the stencil smaller for 128 bit access
2018-07-23 06:12:45 -04:00
e6be7416f4
Use managed memory
2018-06-13 20:14:00 +01:00
26863b6d95
User Managed memory
2018-06-13 20:13:42 +01:00
d4ce7d9905
GPU friendly Stencil needs a view
2018-03-19 07:11:21 -04:00
ad739f042a
Introduce views for passing lattice indexing to accelerators.
2018-03-04 15:56:14 +00:00
bf5fb89aff
Coordinate handling GPU friendly
2018-02-24 22:21:36 +00:00
51eb2c5dfc
Make referencign the stencil and all info required to evaluate the kernel
...
accelerator marked up
2018-02-02 11:37:13 +00:00
79b50feacf
fixme updates
2018-01-29 16:00:40 +00:00
70e276e1ab
parallel_for elimination -> thread_loop
2018-01-28 01:01:14 +00:00
c4f82e072b
_grid becomes private ; use Grid()§
2018-01-27 00:04:12 +00:00
912b50f6fa
Hiding lattice internals
2018-01-26 23:08:45 +00:00
063603b1ea
Warning elimination
2018-01-24 13:12:14 +00:00
5e48b701ec
FOrmatting
2018-01-13 00:11:53 +00:00
7f6bffe5ad
NAMESPACE
2018-01-13 00:11:30 +00:00
0a68470f9a
Simplify comms layers
2018-01-08 11:28:30 +00:00
5918769f97
Subtle Naik term bug updated in Stencil; less on logical && with a function call on right
2017-09-16 12:51:26 +01:00
c289699d9a
updated from cambridge mpi3 shakeout
2017-08-25 11:41:01 +01:00
c3b1263e75
Benchmark prep
2017-08-25 09:25:54 +01:00
a446d95c33
Trying to pass TeamCity and Travis
2017-08-20 01:10:50 +01:00
7d88198387
Merge branch 'develop' into feature/multi-communicator
2017-08-19 13:03:35 -04:00
14d53e1c9e
Threaded MPI calls patches
2017-07-29 13:08:10 -04:00
b5a6e4f1fd
Best option for Xeon cache blocking set
2017-06-30 10:53:22 +01:00
38325ebbc6
Interleave code path; not enabled
2017-06-30 10:23:51 +01:00
54e94360ad
Experimental: Multiple communicators to see if we can avoid thread locks in --enable-comms=mpit
2017-06-24 23:10:24 +01:00
35fa3d1dfd
Merge branch 'master' into feature/scalar_adjointFT
2017-05-12 10:41:39 +01:00
2439999ec8
Warning elimination; drop to -O2 on G++ bad versions
2017-05-06 14:44:49 +01:00
741bc836f6
Exposing support for Ncolours and Ndimensions and JSON input file for the ScalarAction
2017-05-05 17:36:43 +01:00
99220f6531
Fixes and better timing
2017-04-26 17:24:11 -04:00
ab66bac4e6
Think I'm getting on top of the reduced cost exterior precomputed list of links
2017-04-25 08:50:26 +01:00
56277a11c8
Build a list of whats on the surface
2017-04-24 17:06:15 +01:00
3703b718aa
Mark up a table if a given site only receives from itself; including MPI3 splitting info.
2017-04-22 19:28:37 +01:00
736bf3c866
Major rework of stencil. Half precision and MPI3 now working.
2017-04-22 11:33:50 +01:00
e1a2319d01
Simple compressor moved out of cshift into stencil
2017-04-20 13:18:15 +01:00
d2312e9874
Drop compressor entirely from Cshift to only Stencil.
2017-04-20 13:16:55 +01:00
fc4ab9ccd5
Working half precision comms
2017-04-20 11:20:26 +01:00
4a340aa5ca
Massive compressor rework to support reduced precision comms
2017-04-20 09:28:27 +01:00
8ef4300412
spurious .dirstamp files removed
2017-04-10 17:00:22 +01:00
8c8473998d
Average over whole cluster the comm time.
2017-03-21 22:29:51 -04:00
4e7ab3166f
Refactoring header layout
2017-02-22 18:09:33 +00:00
aca7a3ef0a
Optimisation control improvements
2017-02-10 18:22:31 -05:00
c56707e003
useless debug message removed
2016-12-07 08:59:20 +09:00
bb94ddd0eb
Tidy up of mpi3; also some cleaning of the dslash controls.
2016-11-02 08:07:09 +00:00
680645f849
Merge branch 'release/v0.5.0'
2016-06-30 15:15:03 -07:00
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
5e02392f9c
Fixed compilation error for benchmark_dwf
...
Some parts were assuming floating point precision
2016-06-20 12:30:51 +01:00
f78d89bcbe
Update Lebesgue.cc
...
kill verbose
2016-06-03 13:33:42 +01:00
139cc5f1ae
Large change with KNL preparation
2016-06-03 03:24:26 -07:00
