phynexis-deps/armadillo-code
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
2,718 Commits 14 Branches 24 Tags
712596cd5e148bf1bf08dbcd6feb46982357583b
Commit Graph
1999 Commits
Author SHA1 Message Date
conrad 712596cd5e add TODO 2025-08-30 20:37:17 +10:00
conrad 7ee11c94ec add TODO 2025-08-30 20:36:21 +10:00
conrad 8d68fefa48 use threshold size of 30x30 for triggering openmp path 2025-08-30 20:33:38 +10:00
conrad cd1944d9ed simplifications 2025-08-29 16:13:05 +10:00
conrad 877568668d further rework of parallelised matrix mul 2025-08-29 15:48:58 +10:00
conrad 108adbdf97 partial rework on parallelised matrix mul 2025-08-29 14:09:27 +10:00
conrad f76cf14794 separate standard and openmp code paths 2025-08-29 12:30:56 +10:00
Ryan Curtin ced2325184 rudimentary openmp based parallelisation 2025-08-29 01:10:48 +00:00
conrad fe9a74bf75 mark as unstable 2025-08-26 10:58:20 +10:00
conrad 6ddb55c173 patch bump 2025-08-22 15:56:39 +10:00
conrad c029b2a22f use generic function as omp simd seems to have no effect 2025-08-22 11:59:03 +10:00
conrad 8a9f6482bd use pragma omp simd instead of gcc function attribute 2025-08-22 10:58:31 +10:00
conrad bae84c796b allow repcube() to use matrices 2025-08-21 23:49:45 +10:00
conrad 4b0636d0d8 version bummp 2025-08-21 15:14:18 +10:00
conrad cbbf66efc8 add TODO 2025-08-21 14:51:08 +10:00
conrad d59269511f initial version of repcube() 2025-08-21 13:57:18 +10:00
conrad 38795499cc expand debugging to print element type 2025-08-21 12:01:35 +10:00
conrad 8ffcf5383e add TODOs 2025-08-20 18:05:07 +10:00
conrad fdd94a2c81 update group name 2025-08-20 13:59:13 +10:00
conrad 2ca675d242 rename 2025-08-20 13:27:06 +10:00
conrad 9d5cfe2016 use Proxy for compound expressions 2025-08-20 13:17:57 +10:00
conrad 9e1a000e4b check for aliasing only if element types are the same 2025-08-20 12:34:09 +10:00
conrad e10f142788 expand eps() to handle cubes 2025-08-20 12:33:36 +10:00
conrad 4639edc493 return only real for cx inputs 2025-08-20 00:56:11 +10:00
conrad fcd771fda7 sort out the eps() mess 2025-08-20 00:44:22 +10:00
conrad e22c9e7cff update list of functions with optimisation for symmetric matrices 2025-08-19 15:16:06 +10:00
conrad b247c9618e handle corner cases 2025-08-18 16:26:53 +10:00
conrad 17f02679a9 reduce delta 2025-08-18 15:56:57 +10:00
conrad 88fdc045cd add TODOs 2025-08-18 10:43:28 +10:00
conrad 4d2e63c686 fix 2025-08-18 10:39:14 +10:00
Zachary Ng a7a92724d5 add 'valid' convolution mode 2025-08-18 00:35:45 +00:00
conrad 27a9de7f7d simplification 2025-08-16 16:28:14 +10:00
conrad e14cc6fac1 add secondary form of cond() 2025-08-14 19:48:24 +10:00
conrad 1ccc7dbd73 fallback to generic method if symmetric/hermitian optimisation failed 2025-08-14 19:43:08 +10:00
conrad db550a6579 use _OPENMP corresponding to OpenMP 4.0 2025-08-14 15:12:23 +10:00
conrad 5b88b2323a fallback to generic method if symmetric/hermitian optimisation failed 2025-08-14 00:20:59 +10:00
conrad 28ac91cb67 fallback to generic method if symmetric/hermitian optimisation failed 2025-08-14 00:08:33 +10:00
conrad 83452a0627 require OpenMP 4.0 for seq_cst memory order clause 2025-08-13 18:18:33 +10:00
conrad c5cc28a737 change to struct 2025-08-13 18:16:44 +10:00
conrad 9f785eff7e use seq_cst for openmp atomic read/write 2025-08-13 00:19:45 +10:00
conrad cc89707913 update description 2025-08-12 15:34:27 +10:00
conrad efc2bcf316 rearrange 2025-08-11 18:35:11 +10:00
conrad e38cbaab8e handle corner case where tic() is reused 2025-08-11 18:33:09 +10:00
conrad a30ad16ebe rename warm() to unfreeze() 2025-08-11 18:25:11 +10:00
conrad d0da492c26 use type promotion for fp16 version of mat_norm_2() 2025-08-11 17:11:17 +10:00
conrad 3f5d4784fc add freeze() and warm() 2025-08-11 17:03:00 +10:00
conrad 7a15d37a99 simplifications 2025-08-11 14:32:48 +10:00
conrad 1e40f043ff use constexpr 2025-08-11 01:09:38 +10:00
conrad 21f2bf28d2 simplification 2025-08-11 00:51:52 +10:00
conrad 4a5c3d8f27 rearrange 2025-08-11 00:48:59 +10:00