Problem
1. Crossbar switches allow a large number of memory requests to be processed at once, giving excellent performance. Why are they rarely used in practice?
2. A multicomputer with 256 CPUs is organized as a 16 x 16 grid. What is the worst-case delay (in hops) that a message might have to take?
3. Now consider a 256-CPU hypercube. What is the worst-case delay here, again in hops?