Use the DLTlib library to calculate the optimum partitioning for a “query processing” application, where the communication cost is independent of the workload, consisting of a query during...

Use the DLTlib library to calculate the optimum partitioning for a “query processing” application, where the communication cost is independent of the workload, consisting of a query during distribution and the result during the collection. You can assume that b=d=1, the workload consists of L=106items, and the parallel platform is made of 10 compute nodes connected in a single level tree, with pi=(i+1)·0.01sec/ item ∀i∈[0, 9]and l=0.001sec/item. What would happen if the communication were 10 times slower?