[openib-general] RC v/s UD

[dotanb at dev.mellanox.co.il]

One more thing: Did you use the default message size of the tests?
    In ibv_ud_pingpong the default message size is 2K
    In ibv_rc_pingpong the default message size is 4K

so, 2 posts and 2 completions where handled in UD for every 1 post and 1
completion in RC ...

Dotan

> With RC, you can send larger messages and hardware will perform the
> fragmentation.
> BTW, ibv_rc_pingpong/ibv_ud_pingpong are pingpong examples - they do not
> attempt to measure maximum streaming bandwidth.
>
> Look under the perftests directory for some benchmarks:
> e.g. send_bw can measure bandwidth with sends.
>
> Note that with RC, RDMA and SEND bandwidth can also mean different things.
>
>
> Quoting r. john t <johnt1johnt2 at gmail.com>:
> Subject: RC v/s UD
>
> Hi,
>
> I ran "ibv_rc_pingpong" and "ibv_ud_pingpong" utilities and found that RC
> gives
> a BW of 4302 Mbit/sec and UD gives a BW of 2133 Mbit/sec on my setup (2
> hosts
> connected to a switch). So it seems UD is inefficient then RC (gives
> almost
> half the BW as that of RC). Is this expected? Isnt it true that UD
> involves
> less overheads then RC?
>
> Are there ways to get maximum BW (close to RC) from UD?
>
> Regards,
> John T.
>
> _______________________________________________
> openib-general mailing list
> openib-general at openib.org
> http://openib.org/mailman/listinfo/openib-general
>
> To unsubscribe, please visit
> http://openib.org/mailman/listinfo/openib-general
>
> --
> MST
>
> _______________________________________________
> openib-general mailing list
> openib-general at openib.org
> http://openib.org/mailman/listinfo/openib-general
>
> To unsubscribe, please visit
> http://openib.org/mailman/listinfo/openib-general
>