Commit Graph

  • de49a77074 Setting type when gpu sub node is discovered Jack Snyder 2020-08-05 13:34:32 -07:00
  • 3d63f89068
    Merge pull request #364 from badgerious/net-class Sylvain Jeaugey 2020-08-05 12:52:38 -07:00
  • 700c0e0f24 Don't require NIC devices to have specific PCI class Eric Badger 2020-08-03 12:38:07 -07:00
  • 033d799524 2.7.8-1 v2.7.8-1 David Addison 2020-07-23 12:08:08 -07:00
  • 2d8601701d Fix build action order Riatre Foo 2020-07-03 00:07:43 +08:00
  • 1952325569 2.7.6-1 v2.7.6-1 Sylvain Jeaugey 2020-06-23 14:08:49 -07:00
  • 01afd20a77 2.7.5-1 v2.7.5-1 Sylvain Jeaugey 2020-06-22 09:36:20 -07:00
  • 5949d96f36 2.7.3-1 v2.7.3-1 Sylvain Jeaugey 2020-05-12 14:40:18 -07:00
  • f36540f55a Fix crash when only a subset of GPUs are visible within a container. Sylvain Jeaugey 2020-04-17 10:03:14 -07:00
  • 23a9fbb788 Improve robustness of PCI detection Sylvain Jeaugey 2020-04-16 14:27:50 -07:00
  • a783484ab5 Fix wrong variable name "slice" to "chunk" aokomoriuta 2020-01-29 22:05:12 +09:00
  • b5b6c6acdd Fix bug #307 : wrong NIC selection on the reduction tree. Sylvain Jeaugey 2020-04-09 17:14:07 -07:00
  • 533e3702cf
    Merge pull request #314 from NVIDIA/v2.6 Sylvain Jeaugey 2020-03-26 17:31:24 -07:00
  • b221128eca 2.6.4-1 v2.6.4-1 Sylvain Jeaugey 2020-01-16 16:02:42 -08:00
  • 6c61492eba Check return code for Flush operation Rashika Kheria 2020-03-16 18:33:48 -07:00
  • c38f174bd4 Fix Allgather operations above 4G with multiple GPUs per process. Sylvain Jeaugey 2020-02-12 11:04:35 -08:00
  • 3701130b3c 2.5.7-1 v2.5.7-1 Sylvain Jeaugey 2020-01-16 15:40:57 -08:00
  • 44c34e5d10
    Merge pull request #283 from lukeyeager/topo-trim-net-links Sylvain Jeaugey 2020-01-16 15:40:36 -08:00
  • 7a18fe0784 [topology] remove NET links when trimming system Luke Yeager 2020-01-07 13:29:57 -08:00
  • c7ba70ff90 [build] Allow setting CXXFLAGS on the command line Luke Yeager 2020-01-07 13:29:42 -08:00
  • 3899f6e0f2 Fix clang build (#274) Christian Sigg 2019-12-09 18:31:13 +01:00
  • 44b5652617 Merge branch 'master' into HEAD Ke Wen 2019-12-06 18:28:11 -08:00
  • 6bb953d4e6 2.5.6-2 v2.5.6-2 Ke Wen 2019-12-05 18:34:55 -08:00
  • aa15dfb29c Fix clang compilation Sylvain Jeaugey 2019-12-06 09:55:54 -08:00
  • 8c564e9b57 Fix clang build (#271) Christian Sigg 2019-12-06 18:14:55 +01:00
  • 299c554dcc
    2.5.6-1 (#255) v2.5.6-1 Sylvain Jeaugey 2019-11-19 14:57:39 -08:00
  • ccb1298148 Merge branch 'lowintelligence-shm' David Addison 2019-08-14 10:09:53 -07:00
  • fad079a8ae Updated PR#196 to use a common hash function David Addison 2019-08-14 10:08:39 -07:00
  • 01d1836668 Merge branch 'shm' of git://github.com/lowintelligence/nccl into lowintelligence-shm David Addison 2019-08-14 09:45:45 -07:00
  • 7f2b337e70 Make use of SO_REUSEPORT conditional David Addison 2019-08-13 16:32:07 -07:00
  • bfb3921519 Refine RPM package building spec file. Cao Zongyan 2019-07-31 12:21:35 +08:00
  • 4d579e51cc Fix NIC distances for 11+ NICs Ke Wen 2019-07-16 11:28:19 -07:00
  • 920ae57c14 Fix #224: prevent number of IB devices from going out of bound Ke Wen 2019-07-16 08:41:56 -07:00
  • c8c68fb5f7 Size up IPC buffers to multiples of 2MB Avoid potential CUDA error in concurrent communicator initialization Ke Wen 2019-07-12 08:30:05 -07:00
  • 0b192d2299 Add the exact matching modifier support "=" to the NCCL_IB_HCA variable (#236) Hirochika Asai 2019-07-10 06:45:41 +09:00
  • 8e04d80382 Merge branch 'master' into HEAD Ke Wen 2019-06-25 13:39:08 -07:00
  • 7c72dee660 2.4.8-1 v2.4.8-1 Ke Wen 2019-06-25 13:22:47 -07:00
  • 37e4f8729e Fix out-of-bounds read in ncclStrToCpuset (#233) Felix Abecassis 2019-06-21 01:25:08 -07:00
  • 6d8b2421bc Update debian dependencies in README (#228) Rajat Chopra 2019-05-22 21:19:36 -07:00
  • 0ceaec9cee NCCL 2.4.7-1 v2.4.7-1 David Addison 2019-04-16 15:27:06 -07:00
  • 60a586ded9 Allow CUDA runtime library selection (#220) jakirkham 2019-05-07 20:35:14 -04:00
  • 9db4b1d801 Add pkgconfig file (#190) Gustavo Alvarez 2019-04-08 18:16:54 +02:00
  • f40ce73e89 NCCL 2.4.6-1 v2.4.6-1 David Addison 2019-03-14 19:39:20 -07:00
  • 161763aab2 Fix share memory collision in multi-communicator case. Cao Zongyan 2019-03-13 17:13:39 +08:00
  • 14e0cf644b Fix crash during shared memory creation (#185) Rong Ou 2019-03-04 11:42:47 -08:00
  • 1450d42675 2.4.2-1 v2.4.2-1 Sylvain Jeaugey 2018-12-13 15:56:12 -08:00
  • 4861e197fd Fix memory leak in bootstrapRoot() Christian Sigg 2019-01-07 09:34:55 +01:00
  • c244b51ae7 Replace CUDA_VERSION by CUDART_VERSION Sylvain Jeaugey 2018-12-13 15:22:17 -08:00
  • 3e6afef473 Qualify nullptr_t with std:: Christian Sigg 2018-12-13 16:13:53 +01:00
  • 346fc49514 Two temporary workarounds for cuda-clang issues. Christian Sigg 2018-12-13 16:09:12 +01:00
  • d08e9b5279 Change __CUDACC_VER_*__ preprocessor directives to CUDA_VERSION because clang doesn't define the former. Christian Sigg 2018-12-13 15:59:22 +01:00
  • 469b69a5d0 Fix #163 : remove warnings Sylvain Jeaugey 2018-12-11 09:19:16 -08:00
  • 8606cdb8b2 Fix dummy plugin Ke Wen 2018-12-05 17:25:23 -08:00
  • 57368189e1 Remove error logging from a normal path Sylvain Jeaugey 2018-12-04 14:47:41 -08:00
  • 4b39a4cf91 Fix GPU Direct RDMA detection. Sylvain Jeaugey 2018-12-04 14:42:28 -08:00
  • b8a9a32ccb Add NCCL_NET flag to many debug lines. Sylvain Jeaugey 2018-12-04 13:10:19 -08:00
  • cdae05b277 Improve INFO message when external network is not found. Sylvain Jeaugey 2018-12-04 11:57:35 -08:00
  • 5fe2618c0e Fixed some compilation errors when TRACE=1 set David Addison 2018-11-29 13:31:36 -08:00
  • eed8218e17 Rework shared memory code to use SYSCHECK macros. Sylvain Jeaugey 2018-11-05 16:56:14 -08:00
  • 302d538b73 Rework SYSCHECK macros to better handle retries. Sylvain Jeaugey 2018-11-05 16:51:52 -08:00
  • 61b50a63ef Improve net API description Sylvain Jeaugey 2018-11-26 16:23:12 -08:00
  • 98adf2fe11 Make network isend/irecv non blocking Sylvain Jeaugey 2018-11-19 17:43:50 -08:00
  • 0d3a20f96d Add support for external network. Sylvain Jeaugey 2018-11-13 10:37:20 -08:00
  • d7a58cfa58 Generate host-hash for P2P and SHM based on $(readlink /proc/self/ns/uts) + $(readlink /proc/self/ns/mnt) (#156) Alex Sergeev 2018-11-19 17:39:44 -08:00
  • 3c6e25210b
    Generate nccl.h in build instead of src Sylvain Jeaugey 2018-11-09 14:00:41 -08:00
  • 21d9a877be Add official builds download link Ke Wen 2018-11-08 11:22:28 -08:00
  • f7d31919d7
    Add instructions to install packaging toolchain Sylvain Jeaugey 2018-11-05 11:42:33 -08:00
  • bed43524cc Add install target Sylvain Jeaugey 2018-11-05 09:45:46 -08:00
  • b56650c7f5 2.3.7-1 v2.3.7-1 David Addison 2018-10-24 14:44:59 -07:00
  • 3202d6b393 Fix nccl-tests all_reduce_perf path Obihörnchen 2018-10-12 19:08:14 +02:00
  • f93fe9bfd9 2.3.5-5 v2.3.5-5 Sylvain Jeaugey 2018-09-24 16:06:59 -07:00
  • 286916a1a3
    Merge pull request #119 from sclarkson/master Sylvain Jeaugey 2017-11-28 18:41:26 -08:00
  • 680a35c6b7 fix tests on maxwell sclarkson 2017-11-11 19:22:06 -08:00
  • 03d856977e Update README to link to NCCL2 Sylvain Jeaugey 2017-08-04 09:44:37 -07:00
  • 4a33f66e27 Update README to link to NCCL2 part 3 Sylvain Jeaugey 2017-08-04 09:44:09 -07:00
  • d66fb63679 Update README to link to NCCL2 #2 Sylvain Jeaugey 2017-08-04 09:43:29 -07:00
  • 80ae43b443 Update README to link to NCCL2 Sylvain Jeaugey 2017-08-04 09:42:25 -07:00
  • 29a1a916dc Add support for CUDA9 half semantics Sylvain Jeaugey 2017-06-07 09:57:12 -07:00
  • ccfc4567dc Merge pull request #78 from ilya-biryukov/master Sylvain Jeaugey 2017-04-04 09:47:52 -07:00
  • 649f04d077 Added Pascal nvcc flags, bumped version v1.3.4-1 Boris Fomitchev 2017-03-24 11:58:14 -07:00
  • 8241cd7b6e Fix compilation error when compiling with 'clang -x cuda'. Ilya Biryukov 2017-03-16 12:01:11 +01:00
  • 7fef264bfa Bumping version to 1.3.3 Sylvain Jeaugey 2017-03-01 16:43:49 -08:00
  • 8996811936 Only enable peer access for ring neighbors. Nathan Luehr 2017-03-01 15:59:48 -08:00
  • c219a183d0 Fix copy/paste typo in error message Sylvain Jeaugey 2017-02-22 11:56:09 -08:00
  • 8e1d6f9b60 Fix crash in Reduce when non-root ranks have invalid recvbuff Sylvain Jeaugey 2017-02-22 11:55:51 -08:00
  • 024d1e2678 Merge pull request #69 from cwhipkey/master Sylvain Jeaugey 2017-02-08 09:17:50 -08:00
  • 5eab428294 Qualify nullptr_t with std::. Chad Whipkey 2017-02-08 07:06:31 -08:00
  • 2a974f5ca2 Fix 1.3.2 compilation Sylvain Jeaugey 2016-12-08 09:11:43 -08:00
  • 648e9fbb58 Adding missing file Sylvain Jeaugey 2016-12-05 18:06:24 -08:00
  • 34d27771c6 1.3.2 release Sylvain Jeaugey 2016-12-01 15:17:50 -08:00
  • 1093821c33 Replace min BW by average BW in tests Sylvain Jeaugey 2016-12-01 15:16:35 -08:00
  • ddddfba1c0 Merge pull request #54 from peterhj/peterhj-staticlib Sylvain Jeaugey 2016-11-28 09:15:39 -08:00
  • 5765d608cc Add a static library target "staticlib" to the Makefile. Peter Jin 2016-11-23 20:29:39 -08:00
  • c2c515516b Remove irrelevant output from ncclReduce Fortran tests Kyle Fernandes, ne Jacobs 2016-11-20 14:16:53 +00:00
  • 9c18468fe2 Add Copyright header to Fortran bindings source files Kyle Fernandes, ne Jacobs 2016-11-19 12:16:20 +00:00
  • 5f2b32e45b Add Fortran bindings Kyle Fernandes, ne Jacobs 2016-11-17 15:27:20 -08:00
  • 534b9a1697 Bump to 1.3.1 Sylvain Jeaugey 2016-10-13 10:33:05 -07:00
  • b2781d0501 Fix primitives function prototype Sylvain Jeaugey 2016-10-13 10:32:42 -07:00
  • bf7d1514f7 NVML (libwrap) : import the needed definitions Sylvain Jeaugey 2016-10-13 10:28:59 -07:00
  • 8bb06c94be Improved allreduce segmentation for small sizes Sylvain Jeaugey 2016-10-07 12:40:05 -07:00