Sylvain Jeaugey 338f6eaae1 Add support for CUDA9 half semantics ...

[ROCm/rccl commit: 29a1a916dc]

2017-06-14 11:20:24 -07:00

..

all_gather.cu

1.3.2 release

2016-12-01 15:17:50 -08:00

all_reduce.cu

1.3.2 release

2016-12-01 15:17:50 -08:00

broadcast.cu

1.3.2 release

2016-12-01 15:17:50 -08:00

common_coll.h

Fix 1.3.2 compilation

2016-12-08 09:11:43 -08:00

common_kernel.h

Add support for CUDA9 half semantics

2017-06-14 11:20:24 -07:00

copy_kernel.h

Add support for CUDA9 half semantics

2017-06-14 11:20:24 -07:00

core.cu

Only enable peer access for ring neighbors.

2017-03-01 16:42:38 -08:00

core.h

1.3.2 release

2016-12-01 15:17:50 -08:00

enqueue.h

1.3.2 release

2016-12-01 15:17:50 -08:00

libwrap.cu

NVML (libwrap) : import the needed definitions

2016-10-13 10:28:59 -07:00

libwrap.h

NVML (libwrap) : import the needed definitions

2016-10-13 10:28:59 -07:00

nccl.h

Add profiling API

2016-09-22 11:56:51 -07:00

primitives.h

Qualify nullptr_t with std::.

2017-02-08 07:06:31 -08:00

reduce_kernel.h

Heavy code refactoring to remove a lot of code in collectives (~1000 lines).

2016-09-22 11:57:56 -07:00

reduce_scatter.cu

1.3.2 release

2016-12-01 15:17:50 -08:00

reduce.cu

Fix crash in Reduce when non-root ranks have invalid recvbuff

2017-03-01 16:42:38 -08:00