yezhengmao/nccl
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
234 Commits 1 Branch 63 Tags
Commit Graph

234 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sylvain Jeaugey
8bb06c94be Improved allreduce segmentation for small sizes 2016-10-07 12:42:23 -07:00
Sylvain Jeaugey
ca330b110a Add scan tests v1.3.0-1 2016-09-22 11:58:33 -07:00
Sylvain Jeaugey
6c77476cc1 Make tests check for deltas and report bandwidth 2016-09-22 11:58:28 -07:00
Sylvain Jeaugey
cabd6848e4 Heavy code refactoring to remove a lot of code in collectives (~1000 lines). 
Have all collectives use the same args, the same ring, and the same primitives for synchronization between threads with the same pattern.
 2016-09-22 11:57:56 -07:00
Sylvain Jeaugey
e3dbc6110e Add profiling API 2016-09-22 11:56:51 -07:00
Sylvain Jeaugey
1d6715fe20 Fix MPI test path 2016-09-22 11:56:20 -07:00
Sylvain Jeaugey
9ee6189bf9 Merge pull request #41 from jia-kai/master 
Some minor fixes for compile/usage
 2016-09-15 09:45:52 -07:00
Sylvain Jeaugey
939b0a4297 Merge pull request #45 from NVIDIA/cw-update-copyright-year 
Update LICENSE.txt
 2016-08-26 15:44:00 -07:00
Cliff Woolley
234c8c9ef3 Update LICENSE.txt 2016-08-26 15:39:21 -07:00
Sylvain Jeaugey
75bad643bd Updated LICENCE.txt 2016-08-26 15:08:20 -07:00
jiakai
47b0797fe1 pass devlist as const int* rather than int* in ncclCommInitAll 2016-08-19 19:00:14 +08:00
jiakai
ed401cc29b link library with -lrt; otherwise there is undefined reference to shm_open 2016-08-19 18:58:56 +08:00
Sylvain Jeaugey
b3a9e1333d Remove unneeded deb build script 2016-07-27 17:58:00 -07:00
Sylvain Jeaugey
428ec5b2a3 Merge remote-tracking branch 'github/master' into public 2016-07-25 10:53:01 -07:00
Nathan Luehr
55c42ad681 Fixed redundant contexts in multi-process apps 
Change-Id: If787014450fd281304f0c7baf01d25963e40905d
 2016-07-25 10:10:30 -07:00
Sylvain Jeaugey
7a1aa6b563 Improved Deb generation 2016-07-07 16:31:57 +02:00
Sylvain Jeaugey
9ae84f5d6b Fix version number 2016-06-16 17:07:42 -07:00
Sylvain Jeaugey
e51e922924 Add a debug level to NCCL and CUDA versions at init 2016-06-16 17:04:41 -07:00
Sylvain Jeaugey
9fcc523485 Increased version to 1.2.3 2016-06-15 19:18:13 -07:00
Sylvain Jeaugey
67d1ab9106 Packaging : Generate shlibs.local 2016-06-15 19:03:08 -07:00
Sylvain Jeaugey
da6d2009e0 Move deb to build directory 2016-06-15 18:20:10 -07:00
Sylvain Jeaugey
155132d336 Fix make install to use BUILDDIR 2016-06-15 18:20:02 -07:00
Sylvain Jeaugey
08ddfe03d2 Rework debian packaging 2016-06-15 18:18:44 -07:00
Sylvain Jeaugey
5d4716a8a3 Include link to blog post in README.md 2016-06-15 10:54:19 -07:00
Boris Fomitchev
aa8f669a3d Updating for .deb rebuild v1.2.3-1+cuda7.5 2016-06-13 02:01:49 -07:00
Sylvain Jeaugey
d5e507fc7f Only call the CUDA runtime. That may fix #27. 2016-06-07 16:27:51 -07:00
Sylvain Jeaugey
620491a649 Merge remote-tracking branch 'github/master' into HEAD 2016-06-06 14:35:57 -07:00
Sylvain Jeaugey
7edfc57228 Make NCCL collectives work on communicators with only one rank 2016-06-06 14:35:00 -07:00
Sylvain Jeaugey
bd3cf73e6e Changed CURAND generator to work on a wider set of platforms. 2016-06-06 14:34:03 -07:00
Boris Fomitchev
177505b757 Gencodes changed to NV recommended v1.2.1-2+cuda7.5 v1.2.2-1+cuda7.5 2016-06-06 00:06:18 -07:00
Sylvain Jeaugey
9d9d8cd59f Bump to 1.2.2 2016-06-03 17:21:53 -07:00
Sylvain Jeaugey
1657af1567 Better name for GENCODE 2016-06-03 10:25:37 -07:00
Sylvain Jeaugey
acb93d1aed Removing unneeded includes 2016-06-02 17:33:43 -07:00
Sylvain Jeaugey
889ad3d4e6 Makefile improvements 
- Use standard CXX env var
 - Permit redefinition of more env
 - Separate lib from tests
 2016-06-02 15:01:03 -07:00
Boris Fomitchev
93538def65 Merge pull request #22 from borisfom/master 
Fixed version in ChangeLog
v1.2.1-1+cuda7.5 		2016-04-21 18:58:44 -07:00
Boris Fomitchev
e5067b6611 Fixed version in ChangeLog 2016-04-21 16:28:13 -07:00
Boris Fomitchev
0629fb62d7 Merge pull request #21 from borisfom/master 
Fixed install location, new .deb version
 2016-04-21 14:46:41 -07:00
Boris Fomitchev
0177cf3ea4 Fixed install location, new .deb version 2016-04-21 14:10:31 -07:00
Nathan Luehr
658aca1469 Merge pull request #17 from Hopobcn/master 
Enable compilation with specific g++
 2016-04-21 13:25:18 -07:00
Nathan Luehr
03df4c7759 Moved no-as-needed flag to link rule. 
Avoids link errors for tests linked with nvcc.
 2016-04-19 14:51:03 -07:00
Nathan Luehr
0d4f8f4e95 Merge pull request #18 from apaszke/master 
Add --no-as-needed to make sure that cudart library gets linked
 2016-04-19 11:11:39 -07:00
Sylvain Jeaugey
ddd3f2084d Fix readme to reflect the new test paths 2016-04-19 11:09:25 -07:00
Sylvain Jeaugey
dba3ec9428 Fix random deadlock during ncclCommInitRank. 2016-04-19 10:47:27 -07:00
Sylvain Jeaugey
9de361a1b9 Fix MPI test usage 
Only display usage from rank 0 and exit instead of continuing (and seg fault).
 2016-04-19 10:43:38 -07:00
Adam Paszke
c0c959b1be Add --no-as-needed to make sure that cudart library gets liked 2016-04-13 10:04:38 -04:00
Pau Farré
e30bf95989 Enable compilation with old g++ when the default g++ is not supported (+5.0) 2016-04-12 12:49:13 +02:00
Boris Fomitchev
b16cc5d197 Merge pull request #16 from borisfom/master 
Remved Tegra, fixed + format.
v1.2.1-1+cuda7.5-1 		2016-03-17 17:35:04 -07:00
Boris Fomitchev
e6f4a83da6 Removing Tegra 2016-03-17 17:25:27 -07:00
Boris Fomitchev
1a8bae5b2f fixed version format 2016-03-17 17:13:45 -07:00
Boris Fomitchev
e8eb285a59 Merge pull request #15 from borisfom/master 
Fixing version number and compile param for 5.3
 2016-03-17 16:03:05 -07:00