Super-scaling

Eray Ozkural exa at ttnet.net.tr
Wed Feb 19 08:16:29 PST 2003


On Wednesday 19 February 2003 15:30, Simon Hogg wrote:
> Is there a rough rule of thumb which dictates when a program (if ever)
> shows superscaling with number of nodes.  Of course, I would not expect
> this to carry on ad infinitum, but does anyone see superscalar behaviour up
> to, a certain number of nodes.
>
> What would be the conditions for this to occur?

If you mean superlinear speedup there is an obvious explanation for this, your 
parallel algorithm is smarter than the serial algorithm being used. Note 
however that is often not the case and some cache etc. explanation might be 
more valid.

Thanks,

-- 
Eray Ozkural (exa) <erayo at cs.bilkent.edu.tr>
Comp. Sci. Dept., Bilkent University, Ankara
www: http://www.cs.bilkent.edu.tr/~erayo  Malfunction: http://mp3.com/ariza
GPG public key fingerprint: 360C 852F 88B0 A745 F31B  EA0F 7C07 AE16 874D 539C




More information about the Beowulf mailing list