Re: [Gems-users] Multicast in GEMS 1.4

[Marco Solinas <marco.solinas@xxxxxxxxxxxx>]

Hi Niket!
Thanks a lot for your support.
Marco

Niket Agarwal (niketa@xxxxxxxxxxxxx) ha scritto:
> A multicast message in gems 1.4 is not split into multiple unicasts. 
>
> A multicast message travels along as a single message, splitting as and when divergent paths are to be taken for different destinations. 
>
> Niket
>
> ----- Original Message -----
> From: Marco Solinas <marco.solinas@xxxxxxxxxxxx>
> Date: Thursday, September 4, 2008 6:17 am
> Subject: [Gems-users]  Multicast in GEMS 1.4
> To: Gems Users <gems-users@xxxxxxxxxxx>
>
>   
> > Dear List,
> >
> > I was wandering about how GEMS (v 1.4) manages multicast 
> > communication 
> > (when the NetDest of an outgoing message contains more than one 
> > destination node of the system: for example, in case of an 
> > Invalidation 
> > message sent to many -say, N- sharers in a MESI coherence protocol).
> > Is this message split in N unicast messages when leaving the sender 
> > node 
> > output buffer? Or, is it sent as is to ALL the destination? Or, is 
> > it 
> > split when needed, depending on the relative position of 
> > destination 
> > nodes with respect to both the sender and the communication 
> > infrastructure?
> > Thanks a lot for your support.
> > Regards
> > Marco
> >
> > _______________________________________________
> > Gems-users mailing list
> > Gems-users@xxxxxxxxxxx
> > https://lists.cs.wisc.edu/mailman/listinfo/gems-users
> > Use Google to search the GEMS Users mailing list by adding 
> > "site:https://lists.cs.wisc.edu/archive/gems-users/"; to your search.
> >
> >     
> _______________________________________________
> Gems-users mailing list
> Gems-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/gems-users
> Use Google to search the GEMS Users mailing list by adding "site:https://lists.cs.wisc.edu/archive/gems-users/"; to your search.
>
>
>