Tomography is a powerful technique for 3D imaging of the interior of an object. With the growing sizes of typical tomographic data sets, the computational requirements for algorithms in tomography are rapidly increasing. Parallel and distributed-memory methods for tomographic reconstruction are therefore becoming increasingly common. An underexposed aspect is the effect of the data distribution on the performance of distributed-memory reconstruction algorithms. In this work, we introduce a geometric partitioning method, which takes into account the acquisition geometry and aims to minimize the necessary communication between nodes for distributed-memory forward projection and back projection operations. These operations are crucial subroutines for an important class of reconstruction methods. We show that the choice of data distribution has a significant impact on the runtime of these methods. With our novel partitioning method we reduce the communication volume drastically compared to straightforward distributions, by up to 90% for a number of cases, and furthermore we guarantee a specified load balance.

Additional Metadata
Keywords Matrix partitioning, Parallel computing, Sparse matrix, Tomography
Persistent URL dx.doi.org/10.1016/j.parco.2018.12.007
Journal Parallel Computing
Project Real-Time 3D Tomography
Citation
Buurlage, J, Bisseling, R.H, & Batenburg, K.J. (2019). A geometric partitioning method for distributed tomographic reconstruction. Parallel Computing, 81, 104–121. doi:10.1016/j.parco.2018.12.007