Scalable parallel minimum spanning forest computation

Nobari, Sadegh; Cao, Thanh-Tung; Karras, Panagiotis; Bressan, Stéphane

doi:10.1145/2145816.2145842

Cited by 31 publications

(8 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our minimum spanning forest implementation achieves between 17-50x speedup over the implementation running on a single thread across all of our inputs. Obtaining practical parallel algorithms for MSF has been a longstanding goal in the field, and several existing implementations exist [5,23,66,78,90]. We compared our implementation with the union-find based MSF implementation from PBBS [78] and the implementation of Borůvka from [90], which is one of the fastest implementations we are aware of.…”

Section: Methodsmentioning

confidence: 99%

Theoretically Efficient Parallel Graph Algorithms Can Be Fast and Scalable

Dhulipala¹,

Blelloch

Shun³

2021

ACM Trans. Parallel Comput.

View full text Add to dashboard Cite

There has been significant recent interest in parallel graph processing due to the need to quickly analyze the large graphs available today. Many graph codes have been designed for distributed memory or external memory. However, today even the largest publicly-available real-world graph (the Hyperlink Web graph with over 3.5 billion vertices and 128 billion edges) can fit in the memory of a single commodity multicore server. Nevertheless, most experimental work in the literature report results on much smaller graphs, and the ones for the Hyperlink graph use distributed or external memory. Therefore, it is natural to ask whether we can efficiently solve a broad class of graph problems on this graph in memory. This paper shows that theoretically-efficient parallel graph algorithms can scale to the largest publicly-available graphs using a single machine with a terabyte of RAM, processing them in minutes. We give implementations of theoretically-efficient parallel algorithms for 20 important graph problems. We also present the interfaces, optimizations, and graph processing techniques that we used in our implementations, which were crucial in enabling us to process these large graphs quickly. We show that the running times of our implementations outperform existing state-of-the-art implementations on the largest real-world graphs. For many of the problems that we consider, this is the first time they have been solved on graphs at this scale. We have made the implementations developed in this work publicly-available as the Graph Based Benchmark Suite (GBBS).

show abstract

Section: Methodsmentioning

confidence: 99%

Theoretically Efficient Parallel Graph Algorithms Can Be Fast and Scalable

Dhulipala¹,

Blelloch

Shun³

2021

ACM Trans. Parallel Comput.

View full text Add to dashboard Cite

show abstract

“…Our minimum spanning forest implementation achieves between 17-50x speedup over the implementation running on a single thread across all of our inputs. Obtaining practical parallel algorithms for MSF has been a longstanding goal in the field, and several existing implementations exist [9,84,31,106,118]. We compared our implementation with the union-find based MSF implementation from PBBS [106] and the implementation of Borůvka from [118], which is one of the fastest implementations we are aware of.…”

Section: Methodsmentioning

confidence: 99%

Theoretically Efficient Parallel Graph Algorithms Can Be Fast and Scalable

Dhulipala,

Blelloch,

Shun

2018

Preprint

View full text Add to dashboard Cite

There has been significant recent interest in parallel graph processing due to the need to quickly analyze the large graphs available today. Many graph codes have been designed for distributed memory or external memory. However, today even the largest publicly-available real-world graph (the Hyperlink Web graph with over 3.5 billion vertices and 128 billion edges) can fit in the memory of a single commodity multicore server. Nevertheless, most experimental work in the literature report results on much smaller graphs, and the ones for the Hyperlink graph use distributed or external memory. Therefore, it is natural to ask whether we can efficiently solve a broad class of graph problems on this graph in memory.This paper shows that theoretically-efficient parallel graph algorithms can scale to the largest publicly-available graphs using a single machine with a terabyte of RAM, processing them in minutes. We give implementations of theoretically-efficient parallel algorithms for 13 important graph problems. We also present the optimizations and techniques that we used in our implementations, which were crucial in enabling us to process these large graphs quickly. We show that the running times of our implementations outperform existing state-of-the-art implementations on the largest real-world graphs. For many of the problems that we consider, this is the first time they have been solved on graphs at this scale. We provide a publicly-available benchmark suite containing our implementations.

show abstract

“…There have been many works that try to speedup and load balance individual applications, both for distributed and shared memory architectures [4,9,19,20,25,33,37]. The main point is that they present many techniques that are specific to the application under considerations.…”

Section: Load Balancing Graph Applicationsmentioning

confidence: 99%

An Adaptive Load Balancer For Graph Analytical Applications on GPUs

Jatala,

Hoang,

Dathathri

et al. 2019

Preprint

View full text Add to dashboard Cite

Load balancing graph analytics workloads on GPUs is difficult because of the irregular nature of graph applications and the high variability in vertex degrees, particularly in powerlaw graphs. In this paper, we describe a novel load balancing scheme that aims to address this problem. Our scheme is implemented in the IrGL compiler to allow users to generate efficient load-balanced code for GPUs from high-level sequential programs. We evaluated several IrGL-generated load-balanced programs on up to 16 GPUs. Our experiments show that this scheme can achieve an average speed-up of 1.5× on inputs that suffer from severe load imbalance problems when previous state-of-the-art load-balancing schemes are used.

show abstract

Scalable parallel minimum spanning forest computation

Cited by 31 publications

References 20 publications

Theoretically Efficient Parallel Graph Algorithms Can Be Fast and Scalable

Theoretically Efficient Parallel Graph Algorithms Can Be Fast and Scalable

Theoretically Efficient Parallel Graph Algorithms Can Be Fast and Scalable

An Adaptive Load Balancer For Graph Analytical Applications on GPUs

Contact Info

Product

Resources

About