* Added "verifiable", a suite of kernels for generating and verifying reduction
input and output arrays in a bit-precise way.
* Data corruption errors now reported in number of wrong elements instead of max
deviation.
* Use ncclGetLastError.
* Don't run hypercube on non-powers of 2 ranks.
* Fix to hypercube data verification.
* Use "thread local" as the defaut CUDA capture mode.
* Replaced pthread_yield -> sched_yield()
* Bugfix to the cpu-side barrier/allreduce implementations.
Major rework to merge most of the changes from the NCCL internal
tests into the public ones
Added "-m <agg_iters>" operation aggregation option.
Data integrity checking is now much more performant at scale.
Startup times at scale are improved.
Test latency units are now displayed in usec.
Sylvain Jeaugey
John Bachan
David Addison