Hither Hundreds of Processors in a Database Machine

In a recent paper titled “Wither Hundreds of Processors in a Database Machine?”, Agrawal and DeWitt demonstrated that, in order to exploit large numbers of processors in a database machine, the I/O bandwidth of the underlying storage devices must significantly be increased. One way to accomplish this is to use multiple parallel disk units. The main problem with this approach, however, is the lack of a computational model capable of utilizing any significant number of such devices. In this paper we discuss the implementation of a data-driven model (presented at last year’s workshop) on a simulated architecture. It will be shown that this model is capable of exploiting the potential I/O bandwidth of a large number of disk units as well as the computational power of the associated processors.