Computed tomography (CT) image reconstruction algorithms such asconvolution backprojection (CBP) and the Fourier inversion method (FIM)are highly compute intensive applications for today's single processorsystems. Shared memory multiprocessor systems are simple to implement.However the shared bus used to connect the processors to memory is amajor bottleneck in a shared memory multiprocessor system. We havedesigned a hierarchical bus based system (HBBS) which solves the buscongestion problem by hierarchically increasing the number of buses. Thearchitecture is truly expandable and it retains the simplicity of busbased systems. We have implemented the CBP and FIM algorithms on a HBBS,built using commercially available high performance floating pointdigital signal processor (DSP) devices. The HBBS is found to beefficient in executing the computed tomography (CT) image reconstructionalgorithms. The motivation for this work is not to develop newalgorithms, but to evaluate the relative performance of well knownalgorithms when applied to real practical architectures. The resultsshow that an 8-node HBBS executes the CBP and FIM algorithms about 100times faster compared to the execution time of CBP and FIM algorithms onan IBM 340 series Workstation
展开▼
