DPP-PMRF: Rethinking Optimization for a Probabilistic Graphical Model Using Data-Parallel Primitives

TL;DR

This paper introduces a parallel algorithm for optimizing probabilistic graphical models using data-parallel primitives, achieving significant speedups on CPUs and GPUs for image segmentation tasks.

Contribution

The paper presents a novel parallel optimization algorithm based on data-parallel primitives that enhances performance portability across hardware architectures.

Findings

Up to 13X speedup on CPU compared to serial baseline

Up to 44X speedup on GPU compared to serial baseline

Up to 7X speedup over OpenMP-based algorithm on CPU

Abstract

We present a new parallel algorithm for probabilistic graphical model optimization. The algorithm relies on data-parallel primitives (DPPs), which provide portable performance over hardware architecture. We evaluate results on CPUs and GPUs for an image segmentation problem. Compared to a serial baseline, we observe runtime speedups of up to 13X (CPU) and 44X (GPU). We also compare our performance to a reference, OpenMP-based algorithm, and find speedups of up to 7X (CPU).