cuda samples related to cuda::stream does not work

I have come across some problems dealing with cuda streams, that i can not execute nvidia streams samples with excepted results.
code refer to https://github.com/Firststep2014/cuda-sample/tree/master/0_Simple/simpleStreams , results listed as below

-------result-------
Starting Test
memcopy: 5.12
kernel: 0.29
non-streamed: 5.33
4 streams: 5.20

normally, we excepted the steamed result to be much more effective than non-streamed version, so what shall I do next to get more detail.

Device 6: “Tesla V100-PCIE-32GB”
CUDA Driver Version / Runtime Version 10.1 / 10.1

looking forward to your reply, thank you for your kindness