A method of testing p2p Bandwith and latency on nvshmem

Accelerated Computing HPC Compilers nvc, nvc++ and nvfortran
1

Is there any standard method to evaluate the performance on GPU RDMA using Nvshmem ?

I was thinking about an analog of p2pBandwidthLatencyTest but for nvshmem.

2

I’m not aware of anything and in doing a web search only see articles of testing NVShemm performance with specific applications.