Commit Graph
100 Commits
Author SHA1 Message Date
gallo 017cf43381 Add preliminary openacc support, atrip bench not linking 2023-01-11 13:06:59 +01:00
gallo 77e1aaabeb Add bureaucracy for openacc in autotools 2023-01-05 00:06:37 +01:00
gallo 249f1c0b51 Add raven modules for cuda 2023-01-04 15:23:36 +01:00
gallo 1d96800d45 Add support for reading tensors from file in atrip bench 2022-12-06 21:20:03 +01:00
gallo 9087e3af19 Update workflows 2022-12-06 20:58:32 +01:00
gallo 418fd9d389 Add simple cuda bench configuration 2022-12-06 20:57:34 +01:00
gallo 895cd02778 Add some documentation about running the benches 2022-12-06 20:38:57 +01:00
gallo 8efa3d911e Add --max-iterations to main bench 2022-12-06 20:38:38 +01:00
gallo 0fa24404e5 Improve the documentation in the readme for benches building 2022-12-06 14:17:53 +01:00
gallo 8f7d05efda Add Building information and building for sources on GPU 2022-12-06 13:26:44 +01:00
gallo ad542fe856 Add the slicing into the GPU 2022-12-05 21:16:30 +01:00
gallo 658397ebd7 Update in SliceUnion ATRIP_SOURCES_IN_GPU 2022-12-05 17:55:23 +01:00
gallo 26e2f2d109 Add ATRIP_SOURCES_IN_GPU and ATRIP_CUDA_AWARE_MPI defines in configure 2022-12-05 17:49:54 +01:00
gallo 871471aae3 Fix naive-tuples experimentation bench 2022-10-18 16:23:43 +02:00
gallo 65a64f3f8c Test on all pushes 2022-10-08 16:05:40 +02:00
gallo 4f9f09e965 Cleanup flags handling in configure 2022-10-08 16:04:55 +02:00
gallo 6dc943e10a Clean up DatabaseCommunicator 2022-10-08 16:04:37 +02:00
gallo ed347ab0d9 Rename NAIVE_SLOW into ATRIP_NAIVE_SLOW 2022-10-08 16:04:03 +02:00
gallo 8c5c47e208 Add atrip-def m4 macro 2022-10-08 16:03:08 +02:00
gallo 6871372cac Add support for only calculating DGEMM parts ATRIP_ONLY_DGEMM 2022-10-08 16:02:49 +02:00
gallo 452c0fe001 Fix test_main name in automake 2022-10-08 15:59:48 +02:00
gallo b636b89a64 Add configure-benches script 2022-10-08 15:59:25 +02:00
gallo e59d298a01 Rename test_main.cxx to main.cxx 2022-10-08 00:17:57 +02:00
gallo 3ddd507c17 Fix a bug in the building the abc tuples fakely 2022-10-07 12:23:10 +02:00
gallo ae6736fc21 Implement a faster version of naive computation 2022-10-07 03:19:24 +02:00
gallo d5cfe31b12 Add naive tuples scracth file 2022-10-07 01:11:22 +02:00
gallo ddb4574380 Fix DatabaseCommunicator 2022-10-06 12:27:00 +02:00
gallo 118df09128 Add tentative DatabaseCommunicator 2022-10-06 01:10:06 +02:00
gallo 1e391e3749 Update tuples-distribution script 2022-10-06 01:07:53 +02:00
gallo 7734efeb97 Add tuples distribution bench 2022-10-03 17:13:21 +02:00
gallo fa1a29c583 Create an implementation file of the Tuples 2022-10-03 17:11:49 +02:00
gallo 2cbff5c8c9 Add bench utils 2022-10-03 17:11:33 +02:00
gallo 50896e3cd0 Add vendor infrastructure 2022-10-03 17:11:08 +02:00
gallo b11b53aca1 Add warning for not slicing 2022-09-30 14:36:05 +02:00
gallo 5f9725d515 Add forgotten memcpy for SelfSufficient buffers 2022-09-30 14:35:28 +02:00
gallo 1ca31f4929 Fix bench building for cublas-parallel-atrip 2022-09-30 14:34:54 +02:00
gallo 399447131c Exclude clang5 from the nvidia compilation 2022-09-14 14:52:27 +02:00
gallo da2823cf54 Exclude (gcc, cuda) from matrix 2022-09-14 14:43:38 +02:00
gallo 4092f968b7 Tell nvcc to use the correct MPICXX binary
There was a bug related to

  https://gcc.gnu.org/pipermail/gcc-help/2021-June/140490.html

where nvcc was getting other versions of the gcc compiler
and the libstd++ version were mismatched, resulting in linking errors.
2022-09-14 13:28:15 +02:00
gallo a263c0c933 Fix cuda bench 2022-09-13 15:14:45 +02:00
gallo 2895fd25a3 Disable debug in openblas 2022-09-13 15:00:06 +02:00
gallo 8abc516b1f Use --enable-cuda flag in workflows 2022-09-13 14:51:49 +02:00
gallo 4e2de62508 Fix memory consumption in test-cublas-parallel-atrip 2022-09-12 20:19:41 +02:00
gallo 81e89274ab Fix yaml workflows syntax 2022-09-12 20:02:51 +02:00
gallo 5ca94e3238 Try cuda in github ci 2022-09-12 20:00:46 +02:00
gallo 6effcbcdc8 Add ci for cuda branch too 2022-09-12 19:58:18 +02:00
gallo 4277c07cc2 Add memory consumption in bench 2022-09-08 15:44:29 +02:00
gallo 0558148937 Fix small syntactic bug 2022-09-08 15:36:10 +02:00
gallo 49ff3b377c Add a chrono for mpi memcpy in cuda 2022-09-08 15:27:51 +02:00
gallo 00a28c990c Indent more conventionally 2022-09-08 15:27:38 +02:00
gallo 2c5a4620ca Lint and tidy up Equations 2022-09-08 13:51:49 +02:00
gallo 368c5619cc Fix the autotools atrip_cublas 2022-09-08 13:50:08 +02:00
gallo 0b14ac7704 Add bootstrap script 2022-09-06 15:26:28 +02:00
gallo f06cd7f562 Fix the request free problem 2022-08-08 18:28:51 +02:00
gallo 8c04280a65 Fix blas 2022-08-08 18:26:52 +02:00
gallo a30c424781 Fix mkl nix 2022-08-08 18:13:32 +02:00
gallo 0b9ec59599 Fix --with-ctf 2022-08-08 18:13:14 +02:00
gallo 6b641e8581 Update openblas with debug symbols 2022-08-08 15:32:16 +02:00
gallo e03dd77904 Update nvcc compilation in src 2022-07-29 14:08:11 +02:00
gallo b6ca4c8ff2 Update autoconf tests 2022-07-27 18:05:20 +02:00
gallo ad75a3de13 Add __device__ to some functions 2022-07-26 15:12:09 +02:00
gallo 1392e8dc36 Fix some complex pointer errors 2022-07-18 19:27:27 +02:00
gallo e8b733cb1a Fix datafieldtype after doubles 2022-07-18 16:48:57 +02:00
gallo c757c4650c Fix problem with complex numbers 2022-07-14 00:17:24 +02:00
gallo 565fb1dcc8 Autotools changes 2022-07-05 12:45:26 +02:00
gallo 5765ae0dfb Add more implementation files to config 2022-07-05 12:40:25 +02:00
gallo 0f60ffdf0f Initial changes for naive cuda implementation 2022-07-05 12:39:46 +02:00
gallo a4552e2cea Fix the linking of the tests for icc 2022-05-10 12:32:00 +02:00
gallo 40dd346f9a Separate building of ctf from the rest 2022-05-10 12:01:33 +02:00
gallo e621bdda30 Try to factor with shell 2022-05-10 11:55:05 +02:00
gallo bea9c7a75e Tangle sources 2022-05-06 13:58:26 +02:00
gallo d76c33f9e8 Correctly set the world for the application 2022-05-06 13:53:51 +02:00
gallo 2bf8851962 Add Checkpoint functionality 2022-05-06 13:52:33 +02:00
gallo be39eeb776 Add checkpoint readers and writers and tests 2022-04-26 18:42:29 +02:00
gallo 78d67070c0 Fix bench for deleting Vppph 2022-04-05 17:32:21 +02:00
gallo fed19ff52c Implement deleting Vppph 2022-04-05 17:16:11 +02:00
gallo 7d37bd9f8b Update configurations with BLAS_LIBS and so on 2022-03-15 23:38:02 +01:00
gallo b54cdc0573 Silence ctf errors 2022-03-15 23:30:28 +01:00
gallo 62522d0fa6 Add mor extra m4 packages 2022-03-15 23:30:08 +01:00
gallo 1fe8b81bcc Remove check for configure script 2022-03-15 14:57:19 +01:00
gallo fb9fdab703 Update ci for docs with configure 2022-03-15 14:25:31 +01:00
gallo 764d27d67c Docs work 2022-03-14 15:39:13 +01:00
gallo 4d8be00774 Add Documentation to the autotools pipeline 2022-03-14 15:33:10 +01:00
gallo 69611ec0d8 Force stdc++11 2022-03-14 14:06:48 +01:00
gallo 103d7183a0 Again 2022-03-14 13:51:31 +01:00
gallo e16d420b66 Fix shell 2022-03-14 13:49:03 +01:00
gallo fbca9ba447 Now really fixed 2022-03-14 13:44:22 +01:00
gallo bdb8a8be6f Fix env value 2022-03-14 13:43:31 +01:00
gallo 91f9f54015 Update Makefile infrastructure for ctf working build and non-build jobs 2022-03-14 13:41:32 +01:00
gallo 22d4beb33c Update github ci 2022-03-14 13:41:07 +01:00
gallo ef1c50d8a0 Try to compile ctf 2022-03-14 11:43:13 +01:00
gallo 2000de6dfb Update readme with how to cite 2022-03-14 01:10:46 +01:00
gallo a2470048fb Fix math problems with rst 2022-03-14 01:03:39 +01:00
gallo c6eb805078 Now compiling 2022-03-11 14:27:52 +01:00
gallo a7868b49c1 Add more AX_ macros 2022-03-11 12:16:29 +01:00
gallo 07a53aedc5 Add some m4 macros 2022-03-10 22:01:30 +01:00
gallo d51c2a0e4b Initial test for autotools 2022-03-10 22:00:36 +01:00
gallo dae88ba2bd Add autotools in shell.nix 2022-03-10 21:59:18 +01:00
gallo cdb2843a34 Remove hauta 2022-03-10 20:54:00 +01:00
gallo d1c338dcb6 Update doxygen build 2022-03-10 15:48:30 +01:00