A Scalable and Energy Efficient GPU Thread Map for m-Simplex Domains

TL;DR

This paper introduces a new GPU thread mapping technique for m-simplex domains that scales well with dimension, offering significant speedups and energy efficiency improvements over existing methods, especially in higher dimensions.

Contribution

The paper formulates a novel block-space map for m-simplex domains, analyzes its resource usage, and demonstrates its scalability and efficiency through extensive experiments.

Findings

Potential speedup up to 2x for 2-simplices and 6x for 3-simplices.

Achieves 1.2x to 2x speedup for 2-simplices, competitive with state-of-the-art.

Reaches up to 6x speedup for 3-simplices, outperforming existing approaches.

Abstract

This work proposes a new GPU thread map for $m$-simplex domains, that scales its speedup with dimension and is energy efficient compared to other state of the art approaches. The main contributions of this work are i) the formulation of the new block-space map $\mathcal{H}: \mathbb{Z}^m \mapsto \mathbb{Z}^m$ for regular orthogonal simplex domains, which is analyzed in terms of resource usage, and ii) the experimental evaluation in terms of speedup over a bounding box approach and energy efficiency as elements per second per Watt. Results from the analysis show that $\mathcal{H}$ has a potential speedup of up to $2\times$ and $6\times$ for $2$ and $3$-simplices, respectively. Experimental evaluation shows that $\mathcal{H}$ is competitive for $2$-simplices, reaching $1.2\times \sim 2.0\times$ of speedup for different tests, which is on par with the fastest state of the art approaches. For $3$-simplices $\mathcal{H}$ reaches up to $1.3\times \sim 6.0\times$ of speedup making it the fastest of all. The extension of $\mathcal{H}$ to higher dimensional $m$-simplices is feasible and has a potential speedup that scales as $m!$ given a proper selection of parameters $r, \beta$ which are the scaling and replication factors, respectively. In terms of energy consumption, although $\mathcal{H}$ is among the highest in power consumption, it compensates by its short duration, making it one of the most energy efficient approaches. Lastly, further improvements with Tensor and Ray Tracing Cores are analyzed, giving insights to leverage each one of them. The results obtained in this work show that $\mathcal{H}$ is a scalable and energy efficient map that can contribute to the efficiency of GPU applications when they need to process $m$-simplex domains, such as Cellular Automata or PDE simulations.