I remember counting about 52 streams at the test session on 
Supercomputing 05 (if I remember) with about 10% packet loss. Somone may 
have a record better than this and official.

There was about 30 for the Bill Gates Seminar a year or so later.

Agree with Tom: I think we should try a SC Global mass test sometime.

yours Martin

Horwitz, Richard S wrote:
>
> Is there a maximum number of nodes or streams that AG can support?
>
>  
>
>  
>
> Richard Horwitz
>
> Brookhaven National Laboratory
>
> [email protected] <mailto:[email protected]>
>
> (631)344-4134
>

Reply via email to