A current task-based programming paradigms analysis

Abstract

Task-based paradigm models can be an alternative to MPI. The user defines atomic tasks with a defined input and output with the dependencies between them. Then, the runtime can schedule the tasks and data migrations efficiently over all the available cores while reducing the waiting time between tasks. This paper focus on comparing several task-based programming models between themselves using the LU factorization as benchmark. HPX, PaRSEC, Legion and YML+XMP are task-based programming models which schedule data movement and computational tasks on distributed resources allocated to the application. YML+XMP supports parallel and distributed tasks with XscalableMP, a PGAS language. We compared their performances and scalability are compared to ScaLAPACK, an highly optimized library which uses MPI to perform communications between the processes on up to 64 nodes. We performed a block-based LU factorization with the task-based programming model on up to a matrix of size 49512 × 49512. HPX is performing better than PaRSEC, Legion and YML+XMP but not better than ScaLAPACK. YML+XMP has a better scalability than HPX, Legion and PaRSEC. Regent has trouble scaling from 32 nodes to 64 nodes with our algorithm.