Grid Stride Loops in Thrust

Hi,

I would like to use Grid Stride Loop to ensure my parallel kernel launches (each in separate streams) have better guarantee of running simultaneously. Is Thrust for_each using Grid Stride Loops? If it’s not, is there a way to configure it to do so?

Many Thanks,

Ali

Please correct me if I am wrong, It looks like it’s already doing Grid Stride Loop:

[url]https://github.com/thrust/thrust/blob/master/thrust/system/cuda/detail/for_each.inl#L55[/url]