portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-08-21 21:51:54 +01:00
Code Releases Activity

Commit Graph

  • 01231ce824 Stencil fix Peter Boyle 2015-11-29 00:31:02 +00:00
  • ef84d54033 precision set Peter Boyle 2015-11-29 00:30:44 +00:00
  • 41e8038c56 Makefile update Peter Boyle 2015-11-29 00:30:19 +00:00
  • 8a33846095 No compile fix Peter Boyle 2015-11-29 00:29:58 +00:00
  • 54f04ee5c9 Perf event interface was linux specfic and use ifdef to protect Peter Boyle 2015-11-29 00:24:48 +00:00
  • 825875fd48 compile fixes Peter Boyle 2015-11-29 00:24:25 +00:00
  • f8290bfd58 Compile fixes Peter Boyle 2015-11-29 00:24:04 +00:00
  • 967be91692 update merge Azusa Yamaguchi 2015-11-26 09:51:41 +00:00
  • d43034d3ac Merge pull request #21 from aportelli/master azusayamaguchi 2015-11-19 11:45:50 +00:00
  • 06f8ecea04 Merge commit '899ca41cb8c8f47771bfd37cd895cbc2184e5560' Antonin Portelli 2015-11-16 18:16:25 +00:00
  • af19118113 new I/O interface Antonin Portelli 2015-11-16 18:14:37 +00:00
  • e9ff25b06b Small threading change makes a difference on Cori. paboyle 2015-11-07 00:07:05 -08:00
  • 05a7029600 Stencil change paboyle 2015-11-07 00:06:31 -08:00
  • b04b8914fd EXECINFO change paboyle 2015-11-07 00:05:57 -08:00
  • 7522e3f0dd Stencil interface change fix no compile paboyle 2015-11-07 00:05:10 -08:00
  • 1cc0d7b811 Bigger ncall as timing loops got small on cori paboyle 2015-11-07 00:04:40 -08:00
  • 899ca41cb8 Merge branch 'master' of github.com:paboyle/Grid paboyle 2015-11-06 03:50:04 -08:00
  • d29b4c1dee Assembler files paboyle 2015-11-06 03:48:48 -08:00
  • a2ff068e29 Asm and threading for many core paboyle 2015-11-06 03:47:14 -08:00
  • b362f8d27b Threading for many core paboyle 2015-11-06 03:46:41 -08:00
  • 64770d9052 Threading changes for many core and asm calls paboyle 2015-11-06 03:46:21 -08:00
  • 17af18dcab Changes for AVX512 assembler paboyle 2015-11-06 03:45:51 -08:00
  • 28022755ae Stencil class name global change to StencilImpl typedef Peter Boyle 2015-11-06 05:30:17 -06:00
  • 98d8ba6d14 Remove autogen files from CVS Peter Boyle 2015-11-06 05:29:07 -06:00
  • 27813cf518 More timing detail reported Peter Boyle 2015-11-06 05:27:13 -06:00
  • 955b482aaf Partial optimisation of the extraction/merger of simd vecs. Peter Boyle 2015-11-06 05:26:20 -06:00
  • f9b2fce93b Changing whole stencil class to be template and not just single functions Peter Boyle 2015-11-06 05:25:10 -06:00
  • 473fa28a6c Partial optimisation; comms in x-dir for red black dslash will be slow as the checker skipping block strided loops are non threadable. Will need to write a kernel for these instead and drive them with a lookup table to make a look sufficiently simple to thread. Peter Boyle 2015-11-06 05:23:23 -06:00
  • 5d854c869c Stencil interface changes Peter Boyle 2015-11-06 05:22:33 -06:00
  • 880ff88362 Comms optimisation Peter Boyle 2015-11-06 05:22:18 -06:00
  • f85b9ddd97 Remove nonfunctioning lanczos Peter Boyle 2015-11-06 05:21:21 -06:00
  • 4690acc3c8 Don't know why peter committed these as they didn't compile Azusa Yamaguchi 2015-11-06 10:31:48 +00:00
  • 3281745fde Exec info and linux check to stop non-portable code breaking Azusa Yamaguchi 2015-11-06 10:31:24 +00:00
  • c2d96644a0 EXEC INFO check Azusa Yamaguchi 2015-11-06 10:31:05 +00:00
  • 1159de165c Asm option for AVX512 paboyle 2015-11-05 22:04:51 -08:00
  • c4e2202550 First graph class implementation and test Antonin Portelli 2015-11-05 14:28:14 +00:00
  • 16c7993434 Merge branch 'master' of github.com:paboyle/Grid paboyle 2015-11-04 03:32:10 -08:00
  • 6be9716e6f New file paboyle 2015-11-04 03:26:28 -08:00
  • 32762346ad Better run time on KNC paboyle 2015-11-04 03:25:34 -08:00
  • 4a41c885ed Use Linux kernel interface to hardware performance counters. Dead useful. paboyle 2015-11-04 03:24:19 -08:00
  • 0f48658a27 Update minor paboyle 2015-11-04 03:23:46 -08:00
  • 757b31ed42 Threading for KNC mods. paboyle 2015-11-04 03:22:14 -08:00
  • 5aafdd7e1a Inline asm for KNL, KNC, Skylake Xeon paboyle 2015-11-04 03:21:15 -08:00
  • ac7d1f26ad Either blocking or lebesgue curve paboyle 2015-11-04 03:19:16 -08:00
  • 1a8bf938b3 Use either sub-blocking or lebesgue paboyle 2015-11-04 03:18:51 -08:00
  • 63a2993827 Exec info an cache blocking paboyle 2015-11-04 03:16:56 -08:00
  • 4e65ad21ac Adding a routine for AVX512 / IMCI with explicit assembly implementations paboyle 2015-11-04 03:15:08 -08:00
  • dfc1de6f60 Merge branch 'master' of github.com:paboyle/Grid Peter Boyle 2015-11-04 05:14:26 -06:00
  • f87526a04f Make ICC happy Peter Boyle 2015-11-04 05:14:03 -06:00
  • 3b7576ad53 Switch off for now Peter Boyle 2015-11-04 05:13:29 -06:00
  • 9b5d31ffc1 mac , mult routines Lines# with '#' will be ignored, and an empty message aborts the commit. paboyle 2015-11-04 03:10:34 -08:00
  • a38762159c Inline assembly hooks for AVX 512. Better way in some ways than BAGEL to generate assembly. Updated Grid_avx512.h paboyle 2015-11-04 03:09:06 -08:00
  • ffc5dab17f AMD FMA4 support added for Interlagos/BlueWaters Peter Boyle 2015-11-04 04:29:58 -06:00
  • 96608c70d1 chrono causing some problems on Cray systems. Suspend use for now Peter Boyle 2015-11-04 04:28:31 -06:00
  • d35d63b171 Algorithm in Peter Boyle 2015-11-04 04:27:44 -06:00
  • 9183920e8b Added an even odd stencil test, shook out a problem with spread out x-direction. Generalise test to allow different types of "Field" to be used. Peter Boyle 2015-11-04 10:03:04 +00:00
  • 01f286c9fe Better testing for red black cshift which was sufficient to chase down a spread out x-direction problem. Peter Boyle 2015-11-04 10:02:17 +00:00
  • 24044dbc56 Debugged a problem with checkerboarded cshift in the checker dimension which arose only when mpi spread out in the checker dimension. Added a test that trapped and helped debug this Peter Boyle 2015-11-04 10:00:55 +00:00
  • abb23df83f formatting only Peter Boyle 2015-11-04 10:00:27 +00:00
  • 12c5ec813c Useful debug messages (commented out) are included for preservation in case I need to revisit this Peter Boyle 2015-11-04 09:59:27 +00:00
  • 1271508ca2 Bug fix for spread out in x (EO) direction. This is really annoying -- it is very hard to thread the loops with the index recursion on buffer offset in the red-black case. Must think of a good threading solution here. Peter Boyle 2015-11-04 09:57:57 +00:00
  • ec5af35166 EO bug fix when spread out in x-direction Peter Boyle 2015-11-04 09:56:58 +00:00
  • b3d70a3bb2 Ncall change Peter Boyle 2015-11-04 09:55:21 +00:00
  • c26220e9ab EO benchmark as well as non-eo Peter Boyle 2015-11-04 09:54:48 +00:00
  • 0f59356e86 Problem in comms fixed Peter Boyle 2015-11-02 00:00:15 +00:00
  • 538b16610b First commit for measurement software 'Hadrons' Antonin Portelli 2015-10-27 17:33:18 +00:00
  • 8709117aea Log: generalised Logger class to allow separate logs in Grid-based applications Antonin Portelli 2015-10-27 17:31:13 +00:00
  • 1b22ce5720 tests Make.inc fix Antonin Portelli 2015-10-27 10:47:52 +00:00
  • e6b9aa9076 Config.h removed form repository Antonin Portelli 2015-10-27 10:47:07 +00:00
  • d9f2e2e06a Merge pull request #2 from paboyle/master Antonin Portelli 2015-10-19 14:52:52 +01:00
  • 41299da406 files added Peter Boyle 2015-10-09 01:01:46 +02:00
  • 8889af45ca FMA4 added Peter Boyle 2015-10-09 01:00:53 +02:00
  • d4289a33b8 AMD FMA4 addition Peter Boyle 2015-10-09 00:44:20 +02:00
  • 83afb2e26a Poly support for lanczos Peter Boyle 2015-10-09 00:43:21 +02:00
  • 3726fe7481 Bigger vec length Peter Boyle 2015-10-09 00:42:54 +02:00
  • 6d06bd9493 Minor change in commented out code Peter Boyle 2015-10-09 00:42:21 +02:00
  • 6ee23f409e Lanczos addition Peter Boyle 2015-10-09 00:41:00 +02:00
  • 2d95dac6b6 Lanczos untested/partially tested additions. In middle of shake out but at least compiles Peter Boyle 2015-10-09 00:40:25 +02:00
  • 44fecd4d8d Lanczos test Peter Boyle 2015-10-09 00:39:21 +02:00
  • 814c79f38d SIMD improvements for mac and madd use in complex for avx, sse Peter Boyle 2015-10-09 00:38:52 +02:00
  • 1878bf97d0 Babbage fix paboyle 2015-09-30 16:04:01 -07:00
  • 3a478e5f2a No compile babbage fix paboyle 2015-09-30 16:03:05 -07:00
  • a660ce716b No compile babbage fix paboyle 2015-09-30 16:02:44 -07:00
  • f4b6d1dfea NGO stores reenabled paboyle 2015-09-30 16:02:14 -07:00
  • 23813ac798 No compile on babbage fix paboyle 2015-09-30 16:01:28 -07:00
  • af89c40462 Better timing tweaks to give sensible results on 24 threads on Edison dual ivybridge nodes. paboyle 2015-09-28 16:09:04 -07:00
  • 9f4f65cb46 Added a decoupled memory system benchmark to remove thread synch overhead Peter Boyle 2015-09-26 18:23:57 -07:00
  • 64d64d1ab6 Updating to modify non-inlining permute routines and hopefully get better reg use and enhance performance. Peter Boyle 2015-09-25 08:55:04 -07:00
  • 5ef42add2d Changes to remove warnings under icc; disambiguate AVX512 from IMCI correctly and drop swizzles in AVX512. Don't know why these compiled. Peter Boyle 2015-09-23 05:23:45 -07:00
  • 2f38ebc446 Reintroducing the hand unrolled loops Peter Boyle 2015-09-08 17:45:30 +01:00
  • 638d6675ee Tested rms dH is ~ dt^4 numerically, so believe the ForceGradient is correct now. Paranoia makes me want to diddle with the FG step to ensure dt^2 reappears. Peter Boyle 2015-08-31 16:33:20 +01:00
  • 357c6ab46d Reunitarise. Complete the HMC and integrator changes. Peter Boyle 2015-08-31 16:32:04 +01:00
  • 755dca9533 Added ForceGradient integrator. dH dropped so seems to work. Will only believe it is right once I have pulled a dt^4 error scaling plot out. Peter Boyle 2015-08-31 06:23:02 +01:00
  • 29fd004d54 Unified integrator and integrator algorithm into virtual class used as a policy for the HMC. Peter Boyle 2015-08-30 13:39:19 +01:00
  • eed889ea05 Update on todo list Peter Boyle 2015-08-30 12:23:08 +01:00
  • aa52fdadcc Global edit on HMC sector -- making GaugeField a template parameter and preparing to pass integrator, smearing, bc's as policy classes to hmc. Peter Boyle 2015-08-30 12:18:34 +01:00
  • 76d752585b Started a tidy up in the HMC sector. Now comfortable with the two level integrators; to a little figure out what Guido had done & why -- but there is a neat saving of force evaluations across the nesting time boundary making use of linearity of the leapP in dt. Peter Boyle 2015-08-29 17:18:43 +01:00
  • dc814f30da Binary IO file for generic Grid array parallel I/O. Peter Boyle 2015-08-26 13:40:29 +01:00
  • 612957f057 pull in original license. Peter Boyle 2015-08-21 10:19:08 +01:00
  • cea8ac9a22 Credits to orig source where I found the macro tricks. Peter Boyle 2015-08-21 10:14:53 +01:00