Sourav Pati

Followers

Following

Public Views

Graduate student majoring in Computer Science, a two-year degree program at University of Southern California, Los Angeles with a profound interest in applying mathematical modeling and algorithmic principles in coding intelligent computing systems.

less

Jingling Xue

Rachit Rajat

Sanmukh Rao Kuppannagari

Case Western Reserve University

University of Washington

Simon Kahan

University of Washington

Hayden K H So

The University of Hong Kong

Interests

Uploads

Papers by Sourav Pati

Gpop

Proceedings of the 24th Symposium on Principles and Practice of Parallel Programming, 2019

Graph analytics frameworks, typically based on Vertex-centric or Edge-centric paradigms suffer fr... more Graph analytics frameworks, typically based on Vertex-centric or Edge-centric paradigms suffer from poor cache utilization, irregular memory accesses, heavy use of synchronization primitives or theoretical inefficiency, that deteriorate overall performance and scalability. In this paper, we generalize the partition-centric PageRank computation approach [1] to develop a novel Graph Processing Over Partitions (GPOP) framework that enables cache-efficient, work-efficient and scalable implementations of several graph algorithms. For large graphs, we observe that GPOP is upto 19× and 6.1× faster than Ligra and GraphMat, respectively.

Download

GPOP: A cache- and work-efficient framework for Graph Processing Over Partitions

Past decade has seen the development of many shared-memory graph processing frameworks, intended ... more Past decade has seen the development of many shared-memory graph processing frameworks, intended to reduce the effort of developing high performance parallel applications. However many of these frameworks, based on Vertex-centric or Edge-centric paradigms suffer from several issues, such as poor cache utilization, irregular memory accesses, heavy use of synchronization primitives and theoretical inefficiency, that deteriorate overall performance and scalability. Recently, we proposed a cache and memory efficient partition-centric paradigm for computing PageRank. In this paper, we generalize this approach to develop a novel Graph Processing Over Partitions (GPOP) framework that is cache-efficient, scalable and work-efficient. GPOP induces locality in memory accesses by increasing granularity of execution to vertex subsets called 'partitions', thereby dramatically improving the cache performance of a variety of graph algorithms. It achieves high scalability by enabling complet...

Download

Gpop

ACM Transactions on Parallel Computing, Apr 2, 2020

The past decade has seen the development of many shared-memory graph processing frameworks intend... more The past decade has seen the development of many shared-memory graph processing frameworks intended to reduce the effort of developing high-performance parallel applications. However, many of these frameworks, based on Vertex-centric or Edge-centric paradigms suffer from several issues, such as poor cache utilization, irregular memory accesses, heavy use of synchronization primitives, or theoretical inefficiency, that deteriorate over-all performance and scalability. Recently, we proposed a cache and memory-efficient partition-centric paradigm for computing PageRank [26]. In this article, we generalize this approach to develop a novel Graph Processing Over Parts (GPOP) framework that is cache efficient, scalable, and work efficient. GPOP induces locality in memory accesses by increasing granularity of execution to vertex subsets called “parts,” thereby dramatically improving the cache performance of a variety of graph algorithms. It achieves high scalability by enabling completely lock and atomic free computation. GPOP’s built-in analytical performance model enables it to use a hybrid of source and part-centric communication modes in a way that ensures work efficiency each iteration, while simultaneously boosting high bandwidth sequential memory accesses. Finally, the GPOP framework is designed with programmability in mind. It completely abstracts away underlying parallelism and programming model details from the user and provides an easy to program set of APIs with the ability to selectively continue the active vertex set across iterations. Such functionality is useful for many graph algorithms but not intrinsically supported by the current frameworks. We extensively evaluate the performance of GPOP for a variety of graph algorithms, using several large datasets. We observe that GPOP incurs up to 9×, 6.8×, and 5.5× less L2 cache misses compared to Ligra, GraphMat, and Galois, respectively. In terms of execution time, GPOP is up to 19×, 9.3×, and 3.6× faster than Ligra, GraphMat, and Galois, respectively.

Gpop

Proceedings of the 24th Symposium on Principles and Practice of Parallel Programming, 2019

Download

GPOP: A cache- and work-efficient framework for Graph Processing Over Partitions

Download

Gpop

ACM Transactions on Parallel Computing, Apr 2, 2020

Sourav Pati

Related Authors

Uploads

Papers by Sourav Pati

Log In