Commit Graph

75 Commits

Author SHA1 Message Date
Saeed Maleki
56b599b5e7 a bit of api change and clean up on docs 2023-03-24 17:41:04 +00:00
Changho Hwang
274e921009 Minor fixes 2023-03-24 07:28:30 +00:00
Saeed Maleki
c042112b6b perf debug for allgather 2023-03-24 06:49:38 +00:00
Saeed Maleki
777e93ee47 merged with main 2023-03-24 02:35:15 +00:00
Saeed Maleki
56d86472e6 done with allgather_test commenting 2023-03-24 00:22:47 +00:00
Saeed Maleki
58595b1410 Merge pull request #23 from microsoft/saemal/apipush
cleaner allgather_test
2023-03-23 16:17:55 -07:00
Saeed Maleki
86f79ef442 cleaner allgather_test 2023-03-23 23:13:00 +00:00
Madan Musuvathi
e6ee81e4fa fixed the order of remote rank and tag in mscclppConnect API 2023-03-23 21:09:04 +00:00
Madan Musuvathi
72edabe2a6 added GetDevConn api to retrieve a connection from remoteRank and tag 2023-03-23 21:03:30 +00:00
Madan Musuvathi
896539b236 Comm owns all state including devcons 2023-03-22 22:43:32 +00:00
Saeed Maleki
e1cd88ca0b bootstrap test with and without uniq_id 2023-03-22 21:11:27 +00:00
Madan Musuvathi
4c459aa0df allgather_test code cleanup 2023-03-22 20:38:29 +00:00
Madan Musuvathi
261fd7f838 allgather_test code cleanup 2023-03-22 18:50:23 +00:00
Madan Musuvathi
44c6b94747 api version 1 2023-03-22 18:28:30 +00:00
Madan Musuvathi
6ea460bb3a fusing signal with sync 2023-03-22 18:16:42 +00:00
Changho Hwang
9f2eef35d3 Init from a given mscclppUniqueId 2023-03-22 11:25:49 +00:00
Changho Hwang
61511473cb Fix bootstrap_test w/o MPI 2023-03-22 10:03:30 +00:00
Saeed Maleki
483b0c8433 flag is now allocated by the system 2023-03-22 05:14:24 +00:00
Saeed Maleki
0a707d84ec new api works -- single node is not performant 2023-03-22 02:19:49 +00:00
Saeed Maleki
b75f9e6d8a implementing new API 2023-03-22 00:29:10 +00:00
Saeed Maleki
aa1a37ab4d first version 2023-03-21 21:34:19 +00:00
Olli Saarikivi
0cfe2dcffb Add allpairs allreduce test
To support this include separate source and destination offsets in the trigger.
Add functions for getting the rank and world size from a communicator.
2023-03-21 19:00:13 +00:00
Saeed Maleki
e2ee8d80b9 perf fix for multi-node allgather 2023-03-21 06:26:12 +00:00
Saeed Maleki
8b30121240 Merge branch 'main' into chhwang/fix-trigger 2023-03-20 23:12:58 +00:00
Saeed Maleki
7cb2903799 some comment check ins 2023-03-20 21:07:58 +00:00
Saeed Maleki
93afed3e54 new allgather algorithm with both DMA and IB on a single node 2023-03-19 21:53:36 +00:00
Saeed Maleki
8a1ec28ff1 single node allgather works very well 2023-03-19 19:27:17 +00:00
Saeed Maleki
3e8f6758e5 both allgather algorithms 2023-03-19 06:35:40 +00:00
Saeed Maleki
17cbc84a14 both allgather algorithms 2023-03-19 06:35:32 +00:00
Saeed Maleki
a485a7f238 single node works fine -- multinode is problematic 2023-03-19 01:08:05 +00:00
Saeed Maleki
9cc21f70e6 redesigning fifo 2023-03-17 22:51:11 +00:00
Saeed Maleki
73df12358f Merge branch 'main' of https://github.com/microsoft/mscclpp into main 2023-03-17 17:54:17 +00:00
Saeed Maleki
e86df92fa5 fixed a typo in debugging information 2023-03-17 17:52:53 +00:00
Changho Hwang
67dbbd1692 Thread-safe trigger 2023-03-17 09:46:23 +00:00
Saeed Maleki
2061ea91f7 Add allgather_test (#14) 2023-03-17 12:55:20 +08:00
Changho Hwang
aacee9727b trigger wrappers 2023-03-14 09:14:51 +00:00
Saeed Maleki
e000eb9177 some compilation clean up 2023-03-14 05:26:54 +00:00
Saeed Maleki
ab9298d6e0 fixed the bits for trigger 2023-03-13 23:21:27 +00:00
Changho Hwang
e357beef00 One fifo per proxy 2023-03-13 14:19:36 +00:00
Saeed Maleki
ea7134549e vector instructions for trigger 2023-03-13 07:02:26 +00:00
Changho Hwang
1be76d128d 128-bit trigger 2023-03-10 10:49:36 +00:00
Changho Hwang
85d92961a3 Remove MPI dependency 2023-03-10 08:26:38 +00:00
Ubuntu
8f2831330c a few todos + some clean up in the test 2023-03-10 04:34:14 +00:00
Changho Hwang
6ac3c4c90f Relaxed sync 2023-03-09 07:24:09 +00:00
Saeed Maleki
160060ec77 fifo works now 2023-03-08 20:10:09 +00:00
Changho Hwang
1a382a8e1d Fix fifo triggers 2023-03-07 03:31:26 +00:00
Saeed Maleki
3e4c45d73a compiles 2023-03-06 20:36:54 +00:00
Saeed Maleki (saemal)
dced4c4c14 done with the design 2023-03-06 12:26:46 -08:00
Changho Hwang
5ac2ea6e9f IB more fixes 2023-03-06 07:01:03 +00:00
Saeed Maleki
7e4bacf20c works 2023-03-03 23:10:41 +00:00