Commit Graph

50 Commits

Author SHA1 Message Date
Philippe Tillet
006d0f13de Packaging: polished 2015-04-30 00:46:42 -04:00
Philippe Tillet
5ef01f041a Python: Refactored wrapper 2015-04-29 17:48:57 -04:00
Philippe Tillet
cf5028d55b Squashed feature branch:
* Added CUDA support
 * Performance improvements
 * API improvements
 * Added "depth" parameter to GEMM
 * Android cross-compilation
2015-04-29 15:52:21 -04:00
Philippe Tillet
e453031094 More efficient access pattern in the GEMV kernel 2015-02-11 02:06:16 -05:00
Philippe Tillet
7e65601534 fixup 2015-02-08 23:22:48 -05:00
Philippe Tillet
a6d7671831 removing C++11 interface 2015-02-08 23:19:38 -05:00
Philippe Tillet
9c68704f09 Now using a list of event instead of a single one 2015-02-08 00:56:24 -05:00
Philippe Tillet
b768e913c9 Now using events to time autotuning 2015-02-06 22:11:03 -05:00
Philippe Tillet
e214927b16 Better control flow through options 2015-02-05 04:43:50 -05:00
Philippe Tillet
bbf2f0188e Ported to C++11 2015-02-05 04:43:40 -05:00
Philippe Tillet
535706f35a Some renaming; lower overhead in benchmark 2015-02-01 22:28:49 -05:00
Philippe Tillet
f0bb130416 Auto-tuner: Renamed "json_file" to "out" 2015-02-01 21:30:45 -05:00
Philippe Tillet
3f1fa822f8 save 2015-02-01 15:58:05 -05:00
Philippe Tillet
d29f1252ad Clearer array_expression with hopefully lower overhead.
Also removed pyc's
2015-01-31 22:01:48 -05:00
Philippe Tillet
c7665021d1 reducing overhead; reverted custom CL/ header because CL/cl.hpp was buggy 2015-01-28 23:04:19 -05:00
Philippe Tillet
04cec21752 Fixed warnings and compilation for pyatidlas 2015-01-28 19:50:47 -05:00
Philippe Tillet
0dcf4d3617 Better exception handling, lowered CMake requirement ; blas-bench now benchmarks square matrices 2015-01-28 17:08:39 -05:00
Philippe Tillet
c37d8a2a81 Now using system CL include 2015-01-27 16:19:50 -05:00
Philippe Tillet
2a249d26c6 Added postinstall script 2015-01-27 13:06:25 -05:00
Philippe Tillet
909e4b16a0 added preinstall script 2015-01-26 00:56:02 -05:00
Philippe Tillet
6d7f103ab1 Added GEMM benchmark 2015-01-25 18:19:19 -05:00
Philippe Tillet
da9defac04 No more random order in default initialization
Added linking to MKL/CBLAS
2015-01-24 14:51:48 -05:00
Philippe Tillet
9a76be3edc Bugfix in autotuner 2015-01-21 20:08:52 -05:00
Philippe Tillet
0068560bc6 Some cleaning + outer product 2015-01-17 10:49:36 -05:00
Philippe Tillet
f9e5221ec8 removed useless files 2015-01-16 07:38:26 -05:00
Philippe Tillet
69311b7982 Now ATIDLAS is standalone. Everything dynamic.... 2015-01-12 13:24:06 -05:00
Karl Rupp
7c4f8407eb PyATIDLAS: Fixed typo (Bart -> Barts) 2014-11-06 23:12:26 +01:00
Philippe Tillet
6595d20c34 Fixed over-head bug in the auto-tuner (not in the benchmarks) 2014-11-06 16:15:22 -05:00
Karl Rupp
71c72970f6 PyATIDLAS: Added architecture for Maxwell (sm_50) 2014-11-06 13:57:54 +01:00
Philippe Tillet
2be82fc272 Added support for Devastator and Scrapper 2014-11-02 11:28:05 -05:00
Philippe Tillet
f892d4d85d Added pyviennacl as a submodule 2014-11-02 11:05:19 -05:00
Philippe Tillet
6ca682e5eb Added include-guards in viennacl headers generation ; bugfix in autotune.py's operation parsing 2014-11-02 10:31:27 -05:00
Philippe Tillet
0ae87eeff4 Added viennacl-src-path in the UI 2014-11-02 10:05:14 -05:00
Philippe Tillet
45d602bfc8 Better UI 2014-10-31 18:56:33 -04:00
Philippe Tillet
89f3e1d211 Better GA initialization 2014-10-31 18:12:55 -04:00
Philippe Tillet
de48ccc7b1 GEMM now specified in terms of MNK rather than MKN 2014-10-29 23:52:27 -04:00
Philippe Tillet
a8531cac37 Fixed bug in AXPY ; added the possibility to update a .json file 2014-10-29 23:38:22 -04:00
Philippe Tillet
b46f26e54c Auto-tuner: More flexibility 2014-10-29 12:45:20 -04:00
Philippe Tillet
086e51d291 More logical ENUM creation 2014-10-29 17:03:35 +01:00
Philippe Tillet
9803bc8c92 Enhancements of the auto-tuner 2014-10-29 17:01:57 +01:00
Philippe Tillet
199badc689 Cleaned some weird interaction between pyviennacl and pyatidlas 2014-10-28 02:43:42 -04:00
Philippe Tillet
9a5627e20f Prettier command line stuff 2014-10-28 01:10:14 -04:00
Philippe Tillet
1bc7d50258 Now everything is entirely handled through the command line 2014-10-27 22:22:07 -04:00
Philippe Tillet
ba50960b0f Python/Autotune: Moved devices from config.ini to command line argument 2014-10-27 03:28:46 -04:00
Philippe Tillet
11c283590f Replaced cxfreeze with pyinstaller. Works better. 2014-10-16 17:49:17 -04:00
Philippe Tillet
e0f0400a55 Config: Added SONAME for pyatidlas 2014-10-16 09:59:37 -04:00
Philippe Tillet
16c872f38d More packaging ... 2014-10-16 06:57:38 -04:00
Philippe Tillet
d4eb3ed820 Now correctly linking to pyviennacl 2014-10-15 14:21:00 -04:00
Philippe Tillet
30089b7fa6 ATIDLAS's python wrapper is now stand-alone 2014-10-15 05:24:19 -04:00
Philippe Tillet
f91d3b422a Now compiling ATIDLAS 2014-10-14 23:50:35 -04:00