SC13 Research Highlight: Large Graph Processing Without the Overhead

By Dr. Ling Liu and Kisung Lee

November 16, 2013

Many real world information networks consist of millions or billions of vertices representing heterogeneous entities and billions or trillions of edges representing heterogeneous types of relationships among entities.

Image Source: Max Delbrück Center for Molecular Medicine

For example, the crawled Web graph is estimated to have more than 20 billions of pages (vertices) with 160 billions hyperlinks (edges). Facebook user community exceeds 1 billion users (vertices) with more than 140 billion friendship relationships (edges) in 2012.  The billion triple challenges from the Semantic Web community have put forward large collection of RDF datasets with hundreds of millions of vertices and billions of edges.

As the size and variety of information networks continue to grow in many science and engineering domains, graph computations often exceed the processing capacity of conventional hardware, software systems and tools for a number of reasons. First, graph data often exhibits higher data correlations through both direct and indirect edges and such high correlation tends to generate large size of intermediate results during graph computations. When the size of intermediate results exceeds the available memory, the out of memory problem is unavoidable. Second, the graph datasets are growing in volume, variety and velocity. The bigger the size of the graphs gets, the worse the performance will be for most of the graph computations. One open challenge in this space is how to effectively partition a large graph to enable efficient parallel processing of complex graph operations.

One of the papers to be presented at the ACM/IEEE SC13 conference, titled “Efficient data partitioning model for heterogeneous graphs in the Cloud”, proposes a flexible graph partitioning framework, called VB-partitioner. This work is co-authored by the doctorate student Kisung Lee and Prof. Dr. Ling Liu from the school of Computer Science at Georgia Institute of Technology. To make parallel graph computations highly efficient, an important design goal of VB-Partitioner is to devise graph partitioning techniques that can effectively minimize the inter-partition communication overhead and maximize the intra-partition computation capacity (local processing).

Concretely, the first prototype of the VB-Partitioner focuses on efficient processing of graph queries, namely finding all the subgraphs matching a given subgraph pattern. VB-Partitioner partitions a large graph in three steps.

  • First, it constructs three types of Vertex Blocks (in-VBs, out-VBs and bi-VBs) to capture the general graph processing locality.  Each vertex block has an anchor vertex.
  • Second, it constructs three types of k-hop Extended Vertex Blocks (in-EVBs, out-EVBs and bi-EVBs) to distribute vertex blocks with better query locality. Each EVB has one anchor vertex. It achieves query locality by employing controlled edge replication. The setting of k is determined by the radius of frequent query graphs in order to ensure that most frequently requested queries can be processed in parallel at all partitions with minimized inter-partition communication overhead.
  • Third, it partitions a graph by grouping its vertex blocks and EVBs to maximize parallelism in graph processing while ensuring load balance, controlled edge replication and fast grouping.

Four techniques are considered and compared in the context of grouping and placement of VBs and EVBs to partitions: random grouping, hash-based grouping, min-cut based grouping and high degree vertex-based grouping.  As an integral part of the VB-Partitioner, a data partition-aware query partitioning model is also developed to handle the cases where the radius of a query is larger than k. The experimental results reported in the paper demonstrate the effectiveness of VB-Partitioner in terms of query processing efficiency, data loading time and partition distribution balance.

Graph computations can be broadly classified into two categories, graph queries that find matching subgraphs of a given pattern and iterative graph algorithms that find clusters, orderings, paths or correlation patterns. The former targets at subgraph matching problems over large static graphs and the later targets at graph inference kernels that traverse the graph by iteratively updating the weight of vertices or edges, such as PageRank, shortest path algorithms, spanning tree algorithms, topological sort, and so forth.

Although the first generation of the VB-Partitioner is tailored primarily for efficient parallel processing of graph queries, the ongoing work on VB-Partitioner includes exploring the feasibility and effectiveness of VB-Partitioner in the context of iterative graph algorithms. For example, to minimize inter-partition communications and maximize parallelism in graph computation, it is crucial to optimize the shared memory by minimizing parallel overhead of synchronization barriers. It is equally important to optimize the distributed memory by bounding message buffer sizes, bundling messages, overlapping communication with computation to amortize the overhead of barriers.

image1
Image Source: Giot et al., “A Protein Interaction Map of Drosophila melanogaster”, Science 302, 1722-1736, 2003.”

In addition to exploring parallel computation opportunities through graph partitioning using multi-threads, multi-cores, multiple workers, one can also exploit and combine with other performance optimization techniques to scale large graph analytics. Example techniques include

  • Compression to provide compact storage and in-memory processing,
  • Data placements on disk and in memory to balance computation with storage, and to maximize sequential access and minimize random access,
  • Indexing at vertex and/or edge level to utilize sequential access and minimize unnecessary random access,
  • Caching at vertex, edge or query level to gain performance for repeated access.

Please come hear more on Tuesday, November 19, 2013 10:30AM – 11:00AM (Location: Room 205/207)

http://sc13.supercomputing.org/schedule/event_detail.php?evid=pap708

About the Authors

LingLing Liu is a Professor in the School of Computer Science at Georgia Institute of Technology. She directs the research programs in Distributed Data Intensive Systems Lab (DiSL), examining various aspects of large scale data intensive systems. Prof. Ling Liu is an internationally recognized expert in the areas of Cloud computing, Distributed Computing, Big Data technologies, Database systems and Service oriented computing. Prof. Liu is a recipient of IEEE Computer Society Technical Achievement Award in 2012. Currently Prof. Liu is the editor in chief of IEEE Transactions on Service Computing, and serves on the editorial board of half dozen international journals, including Journal of Parallel and Distributed Computing (JPDC), ACM Transactions on Internet Technology (TOIT), ACM Transactions on Web (TWEB). Dr. Liu’s current research is primarily sponsored by NSF, IBM, and Intel.

 

luiKisung Lee is a Ph.D student in the School of Computer Science at Georgia Tech. He received his BS and MS degree in computer science from KAIST in 2005 and 2007 respectively. He had worked for ETRI as a researcher from 2007 to 2010. He is conducting research in distributed and parallel processing of big data in the Cloud, mobile computing and social network analysis.

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industy updates delivered to you every week!

With New Owner and New Roadmap, an Independent Omni-Path Is Staging a Comeback

July 23, 2021

Put on a shelf by Intel in 2019, Omni-Path faced a uncertain future, but under new custodian Cornelis Networks, OmniPath is looking to make a comeback as an independent high-performance interconnect solution. A "significant refresh" – called Omni-Path Express – is coming later this year according to the company. Cornelis Networks formed last September as a spinout of Intel's Omni-Path division. Read more…

PEARC21 Panel Reviews Eight New NSF-Funded HPC Systems Debuting in 2021

July 23, 2021

Over the past few years, the NSF has funded a number of HPC systems to further supply the open research community with computational resources to meet that community’s changing and expanding needs. A review of these systems at the PEARC21 conference (July 19-22) highlighted... Read more…

Chameleon’s HPC Testbed Sharpens Its Edge, Presses ‘Replay’

July 22, 2021

“One way of saying what I do for a living is to say that I develop scientific instruments,” said Kate Keahey, a senior fellow at the University of Chicago and a computer scientist at Argonne National Laboratory, as s Read more…

PEARC21 Plenary Session: AI for Innovative Social Work

July 21, 2021

AI analysis of social media poses a double-edged sword for social work and addressing the needs of at-risk youths, said Desmond Upton Patton, senior associate dean, Innovation and Academic Affairs, Columbia University. S Read more…

Summer Reading: “High-Performance Computing Is at an Inflection Point”

July 21, 2021

At last month’s 11th International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies (HEART), a group of researchers led by Martin Schulz of the Leibniz Supercomputing Center (Munich) presented a “position paper” in which they argue HPC architectural landscape... Read more…

AWS Solution Channel

Accelerate innovation in healthcare and life sciences with AWS HPC

With Amazon Web Services, researchers can access purpose-built HPC tools and services along with scientific and technical expertise to accelerate the pace of discovery. Whether you are sequencing the human genome, using AI/ML for disease detection or running molecular dynamics simulations to develop lifesaving drugs, AWS has the infrastructure you need to run your HPC workloads. Read more…

PEARC21 Panel: Wafer-Scale-Engine Technology Accelerates Machine Learning, HPC

July 21, 2021

Early use of Cerebras’ CS-1 server and wafer-scale engine (WSE) has demonstrated promising acceleration of machine-learning algorithms, according to participants in the Scientific Research Enabled by CS-1 Systems panel Read more…

With New Owner and New Roadmap, an Independent Omni-Path Is Staging a Comeback

July 23, 2021

Put on a shelf by Intel in 2019, Omni-Path faced a uncertain future, but under new custodian Cornelis Networks, OmniPath is looking to make a comeback as an independent high-performance interconnect solution. A "significant refresh" – called Omni-Path Express – is coming later this year according to the company. Cornelis Networks formed last September as a spinout of Intel's Omni-Path division. Read more…

Chameleon’s HPC Testbed Sharpens Its Edge, Presses ‘Replay’

July 22, 2021

“One way of saying what I do for a living is to say that I develop scientific instruments,” said Kate Keahey, a senior fellow at the University of Chicago a Read more…

Summer Reading: “High-Performance Computing Is at an Inflection Point”

July 21, 2021

At last month’s 11th International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies (HEART), a group of researchers led by Martin Schulz of the Leibniz Supercomputing Center (Munich) presented a “position paper” in which they argue HPC architectural landscape... Read more…

PEARC21 Panel: Wafer-Scale-Engine Technology Accelerates Machine Learning, HPC

July 21, 2021

Early use of Cerebras’ CS-1 server and wafer-scale engine (WSE) has demonstrated promising acceleration of machine-learning algorithms, according to participa Read more…

15 Years Later, the Green500 Continues Its Push for Energy Efficiency as a First-Order Concern in HPC

July 15, 2021

The Green500 list, which ranks the most energy-efficient supercomputers in the world, has virtually always faced an uphill battle. As Wu Feng – custodian of the Green500 list and an associate professor at Virginia Tech – tells it, “noone" cared about energy efficiency in the early 2000s, when the seeds... Read more…

Frontier to Meet 20MW Exascale Power Target Set by DARPA in 2008

July 14, 2021

After more than a decade of planning, the United States’ first exascale computer, Frontier, is set to arrive at Oak Ridge National Laboratory (ORNL) later this year. Crossing this “1,000x” horizon required overcoming four major challenges: power demand, reliability, extreme parallelism and data movement. Read more…

Quantum Roundup: IBM, Rigetti, Phasecraft, Oxford QC, China, and More

July 13, 2021

IBM yesterday announced a proof for a quantum ML algorithm. A week ago, it unveiled a new topology for its quantum processors. Last Friday, the Technical Univer Read more…

ExaWind Prepares for New Architectures, Bigger Simulations

July 10, 2021

The ExaWind project describes itself in terms of terms like wake formation, turbine-turbine interaction and blade-boundary-layer dynamics, but the pitch to the Read more…

AMD Chipmaker TSMC to Use AMD Chips for Chipmaking

May 8, 2021

TSMC has tapped AMD to support its major manufacturing and R&D workloads. AMD will provide its Epyc Rome 7702P CPUs – with 64 cores operating at a base cl Read more…

Intel Launches 10nm ‘Ice Lake’ Datacenter CPU with Up to 40 Cores

April 6, 2021

The wait is over. Today Intel officially launched its 10nm datacenter CPU, the third-generation Intel Xeon Scalable processor, codenamed Ice Lake. With up to 40 Read more…

Berkeley Lab Debuts Perlmutter, World’s Fastest AI Supercomputer

May 27, 2021

A ribbon-cutting ceremony held virtually at Berkeley Lab's National Energy Research Scientific Computing Center (NERSC) today marked the official launch of Perlmutter – aka NERSC-9 – the GPU-accelerated supercomputer built by HPE in partnership with Nvidia and AMD. Read more…

Ahead of ‘Dojo,’ Tesla Reveals Its Massive Precursor Supercomputer

June 22, 2021

In spring 2019, Tesla made cryptic reference to a project called Dojo, a “super-powerful training computer” for video data processing. Then, in summer 2020, Tesla CEO Elon Musk tweeted: “Tesla is developing a [neural network] training computer called Dojo to process truly vast amounts of video data. It’s a beast! … A truly useful exaflop at de facto FP32.” Read more…

Google Launches TPU v4 AI Chips

May 20, 2021

Google CEO Sundar Pichai spoke for only one minute and 42 seconds about the company’s latest TPU v4 Tensor Processing Units during his keynote at the Google I Read more…

CentOS Replacement Rocky Linux Is Now in GA and Under Independent Control

June 21, 2021

The Rocky Enterprise Software Foundation (RESF) is announcing the general availability of Rocky Linux, release 8.4, designed as a drop-in replacement for the soon-to-be discontinued CentOS. The GA release is launching six-and-a-half months after Red Hat deprecated its support for the widely popular, free CentOS server operating system. The Rocky Linux development effort... Read more…

CERN Is Betting Big on Exascale

April 1, 2021

The European Organization for Nuclear Research (CERN) involves 23 countries, 15,000 researchers, billions of dollars a year, and the biggest machine in the worl Read more…

Iran Gains HPC Capabilities with Launch of ‘Simorgh’ Supercomputer

May 18, 2021

Iran is said to be developing domestic supercomputing technology to advance the processing of scientific, economic, political and military data, and to strengthen the nation’s position in the age of AI and big data. On Sunday, Iran unveiled the Simorgh supercomputer, which will deliver.... Read more…

Leading Solution Providers

Contributors

HPE Launches Storage Line Loaded with IBM’s Spectrum Scale File System

April 6, 2021

HPE today launched a new family of storage solutions bundled with IBM’s Spectrum Scale Erasure Code Edition parallel file system (description below) and featu Read more…

Julia Update: Adoption Keeps Climbing; Is It a Python Challenger?

January 13, 2021

The rapid adoption of Julia, the open source, high level programing language with roots at MIT, shows no sign of slowing according to data from Julialang.org. I Read more…

10nm, 7nm, 5nm…. Should the Chip Nanometer Metric Be Replaced?

June 1, 2020

The biggest cool factor in server chips is the nanometer. AMD beating Intel to a CPU built on a 7nm process node* – with 5nm and 3nm on the way – has been i Read more…

GTC21: Nvidia Launches cuQuantum; Dips a Toe in Quantum Computing

April 13, 2021

Yesterday Nvidia officially dipped a toe into quantum computing with the launch of cuQuantum SDK, a development platform for simulating quantum circuits on GPU-accelerated systems. As Nvidia CEO Jensen Huang emphasized in his keynote, Nvidia doesn’t plan to build... Read more…

Microsoft to Provide World’s Most Powerful Weather & Climate Supercomputer for UK’s Met Office

April 22, 2021

More than 14 months ago, the UK government announced plans to invest £1.2 billion ($1.56 billion) into weather and climate supercomputing, including procuremen Read more…

Q&A with Jim Keller, CTO of Tenstorrent, and an HPCwire Person to Watch in 2021

April 22, 2021

As part of our HPCwire Person to Watch series, we are happy to present our interview with Jim Keller, president and chief technology officer of Tenstorrent. One of the top chip architects of our time, Keller has had an impactful career. Read more…

Quantum Roundup: IBM, Rigetti, Phasecraft, Oxford QC, China, and More

July 13, 2021

IBM yesterday announced a proof for a quantum ML algorithm. A week ago, it unveiled a new topology for its quantum processors. Last Friday, the Technical Univer Read more…

Senate Debate on Bill to Remake NSF – the Endless Frontier Act – Begins

May 18, 2021

The U.S. Senate today opened floor debate on the Endless Frontier Act which seeks to remake and expand the National Science Foundation by creating a technology Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire