large clusters and topologies

[Alan Ward]

This discussion we've been having about > 100 node clusters
makes me wonder about two points:

- I suppose Ethernet is not all too efficient with this
size, even with switches instead of hubs. What's the largest
practical size for an Ethernet 'wulf? Any alternatives to
Myrinet upwards of that?

- Which type of algorithm can take direct advantage of a
cube, 
toroid, ... topology? I would think even a "dumb" job like 
matrix multiplication might benefit from a 2D lattice
(instead
of the typical all-to-all unique segment). Comments?


Best regards,
Alan Ward