Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/marian-nmt/nccl.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2018-10-252.3.7-1v2.3.7-1David Addison
Improved LL tuning for multi-node jobs. Improved bootstrap for large job scaling. Fixed a hang during bootstrap due to socket reuse. Added operation name to the COLL INFO logging.
2018-10-14Fix nccl-tests all_reduce_perf pathObihörnchen
It's `all_reduce_perf` not `allreduce_perf`
2018-09-262.3.5-5v2.3.5-5Sylvain Jeaugey
Add support for inter-node communication using sockets and InfiniBand/RoCE. Improve latency. Add support for aggregation. Improve LL/regular tuning. Remove tests as those are now at github.com/nvidia/nccl-tests .
2017-11-29Merge pull request #119 from sclarkson/masterSylvain Jeaugey
Fix tests: call cudaHostUnregister on the host pointer instead of the device pointer.
2017-11-12fix tests on maxwellsclarkson
2017-08-04Update README to link to NCCL2Sylvain Jeaugey
2017-08-04Update README to link to NCCL2 part 3Sylvain Jeaugey
2017-08-04Update README to link to NCCL2 #2Sylvain Jeaugey
2017-08-04Update README to link to NCCL2Sylvain Jeaugey
2017-06-14Add support for CUDA9 half semanticsSylvain Jeaugey
2017-04-04Merge pull request #78 from ilya-biryukov/masterSylvain Jeaugey
Fix compilation error when compiling with 'clang -x cuda'.
2017-03-24Added Pascal nvcc flags, bumped versionv1.3.4-1Boris Fomitchev
2017-03-16Fix compilation error when compiling with 'clang -x cuda'.Ilya Biryukov
Functions vFetch and vStore are not found by ADL with clang, so they need to be declared before usage in ReduceCopy.
2017-03-02Bumping version to 1.3.3Sylvain Jeaugey
2017-03-02Only enable peer access for ring neighbors.Nathan Luehr
This enables support for systems with more than 9 GPUs attached to a single PCIe root complex.
2017-03-02Fix copy/paste typo in error messageSylvain Jeaugey
2017-03-02Fix crash in Reduce when non-root ranks have invalid recvbuffSylvain Jeaugey
2017-02-08Merge pull request #69 from cwhipkey/masterSylvain Jeaugey
Qualify nullptr_t with std::
2017-02-08Qualify nullptr_t with std::.Chad Whipkey
2016-12-08Fix 1.3.2 compilationSylvain Jeaugey
2016-12-06Adding missing fileSylvain Jeaugey
2016-12-021.3.2 releaseSylvain Jeaugey
Broadcast tuning Better checking of inputs Copy/reduce code simplification
2016-12-02Replace min BW by average BW in testsSylvain Jeaugey
2016-11-28Merge pull request #54 from peterhj/peterhj-staticlibSylvain Jeaugey
Add a static library target "staticlib" to the Makefile.
2016-11-24Add a static library target "staticlib" to the Makefile.Peter Jin
Rename the static library "libnccl_static.a" to disambiguate from the dynamic libraries.
2016-11-21Remove irrelevant output from ncclReduce Fortran testsKyle Fernandes, ne Jacobs
2016-11-21Add Copyright header to Fortran bindings source filesKyle Fernandes, ne Jacobs
2016-11-18Add Fortran bindingsKyle Fernandes, ne Jacobs
2016-10-13Bump to 1.3.1Sylvain Jeaugey
2016-10-13Fix primitives function prototypeSylvain Jeaugey
2016-10-13NVML (libwrap) : import the needed definitionsSylvain Jeaugey
2016-10-07Improved allreduce segmentation for small sizesSylvain Jeaugey
2016-09-22Add scan testsv1.3.0-1Sylvain Jeaugey
2016-09-22Make tests check for deltas and report bandwidthSylvain Jeaugey
2016-09-22Heavy code refactoring to remove a lot of code in collectives (~1000 lines).Sylvain Jeaugey
Have all collectives use the same args, the same ring, and the same primitives for synchronization between threads with the same pattern.
2016-09-22Add profiling APISylvain Jeaugey
2016-09-22Fix MPI test pathSylvain Jeaugey
2016-09-15Merge pull request #41 from jia-kai/masterSylvain Jeaugey
Some minor fixes for compile/usage
2016-08-27Merge pull request #45 from NVIDIA/cw-update-copyright-yearSylvain Jeaugey
Update LICENSE.txt
2016-08-27Update LICENSE.txtCliff Woolley
2016-08-27Updated LICENCE.txtSylvain Jeaugey
2016-08-19pass devlist as const int* rather than int* in ncclCommInitAlljiakai
2016-08-19link library with -lrt; otherwise there is undefined reference to shm_openjiakai
2016-07-28Remove unneeded deb build scriptSylvain Jeaugey
2016-07-25Merge remote-tracking branch 'github/master' into publicSylvain Jeaugey
2016-07-25Fixed redundant contexts in multi-process appsNathan Luehr
Change-Id: If787014450fd281304f0c7baf01d25963e40905d
2016-07-07Improved Deb generationSylvain Jeaugey
2016-06-17Fix version numberSylvain Jeaugey
2016-06-17Add a debug level to NCCL and CUDA versions at initSylvain Jeaugey
2016-06-16Increased version to 1.2.3Sylvain Jeaugey
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-26 16:55:31 +0300