19b527e83f
Better extract merge for GPU. Let the SIMD header files define the pointer type for
...
access. GPU redirects through builtin float2, double2 for complex
2018-07-05 07:05:13 -04:00
09cd46d337
Lane by Lane operation
2018-05-12 17:59:35 -04:00
90a2efb9b3
Hit an annoying strict alias optimisation in GCC 4.9 through 6.3
...
Chris K was correct. It appears that an additional memcpy (UGHHH) is enough
to suppress the compiler
2018-03-07 07:27:26 -08:00
4d53703c67
Scalar type differeing allowed, eg. precisoin change
2018-03-05 11:39:52 +00:00
e5ea04ee0c
Need to support precision change, and real replication in multiple simd lanes
2018-03-04 15:53:04 +00:00
7574c18cef
Massive clean up extract merge.
...
Simpler and GPU friendly
2018-02-24 22:21:08 +00:00
8e99264f40
Accelerator mark up of entire tensore space for offload
2018-01-24 13:27:30 +00:00
c037244874
Tensor reformatted with NAMESPACE too
2018-01-13 00:31:02 +00:00
3cbe974eb4
Layout
2016-10-20 16:55:21 +01:00
85ed8175cb
Implemented mixed precision CG. Fixed filelist to exclude lib/Old directory and include Config.h.
2016-07-06 15:57:04 -04:00
8fd8bc25e9
simd 5th dim with rotation
2016-04-19 15:39:00 -07:00
c9fadf97a5
Simplify the compressor interface again.
2016-02-17 18:16:45 -06:00
c650bb3f3d
Very small merge speed up.
2016-02-16 18:41:53 -06:00
fc6ad65751
Pushed the overlap comms tweaks
2016-01-11 06:34:22 -08:00
aae8bf31a7
Global edit adding copyright and license info to every source file.
2016-01-02 14:51:32 +00:00
145a295231
Bug fix for stencil with large shifts (3+), would be important to naik term for example but did not
...
impact Wilson based nearest neighbour stencils.
2015-12-30 19:29:48 +00:00
955b482aaf
Partial optimisation of the extraction/merger of simd vecs.
2015-11-06 05:26:20 -06:00
d1afebf71e
Sizable improvement in multigrid for unsquared.
...
6000 matmuls CG unprec
2000 matmuls CG prec (4000 eo muls)
1050 matmuls PGCR on 16^3 x 32 x 8 m=.01
Substantial effort on timing and logging infrastructure
2015-07-24 01:31:13 +09:00
6e5db0b1da
Corrected bug in integer multiplications for SSE4 and AVX2
...
Merge remote-tracking branch 'upstream/master'
Conflicts:
tests/Make.inc
2015-06-16 23:34:45 +09:00
ef97692622
Handle case of simd_layout not filling whole vector.
...
Useful if real complex live on same grid
2015-06-14 00:55:21 +01:00
1d0df449e8
Reorganise of file naming
2015-06-03 12:47:05 +01:00
