Skip to content

Tags: Rongnian/nccl

Tags

v2.3.7-1

Toggle v2.3.7-1's commit message 
2.3.7-1

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.

v2.3.5-5

Toggle v2.3.5-5's commit message 
2.3.5-5

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 .

v1.3.4-1

Toggle v1.3.4-1's commit message 
Added Pascal nvcc flags, bumped version

v1.3.0-1

Toggle v1.3.0-1's commit message 
Add scan tests

v1.2.3-1+cuda8.0

Toggle v1.2.3-1+cuda8.0's commit message 
Preparing for 1.2.3 rebuild

v1.2.3-1+cuda7.5

Toggle v1.2.3-1+cuda7.5's commit message 
Updating for .deb rebuild

v1.2.2-1+cuda8.0

Toggle v1.2.2-1+cuda8.0's commit message 
Gencodes changed to NV recommended

v1.2.2-1+cuda7.5

Toggle v1.2.2-1+cuda7.5's commit message 
Gencodes changed to NV recommended

v1.2.1-2+cuda7.5

Toggle v1.2.1-2+cuda7.5's commit message 
Gencodes changed to NV recommended

v1.2.1-1+cuda7.5

Toggle v1.2.1-1+cuda7.5's commit message 
Merge pull request NVIDIA#22 from borisfom/master

Fixed version in ChangeLog