Jump to content Australia-English
HP.com Australia home Products and Services Support and Drivers Solutions How to Buy
» Contact HP
HP.com Australia home
HPC  >  News

Grid myths debunked

» 

High Performance Computing

Products & Solutions

» Systems
» Clusters
» Storage
» Visualization
» Software
» Grid computing
» Services and Support
» Solutions
» Power and Cooling

Resources

» News & Promotions
» Subscribe to newsletter
» HPC Library
» HPC Podcasts
» HPC events
» Benchmarks
» How to Buy

Learn more about

» Customer successes
» HP Collaboration & Competency Network
» HP-CAST User Group meetings
» Catalysts for Innovation Program
» HP Labs
» HP Partners
» Explore the HP Alternative program
» Contact HP
Content starts here

A cursory glance of recent computer tabloid headlines will tell you grids are the next big thing to hit IT. From promises of utility-style computing to talk of mega-grids, just about every vendor has a definition or technology to sell.

But before you get caught up in the hype, there's something to keep in mind: a grid is not a product, it's an evolution, says Sara Murphy, HP marketing manager for Grids.

“Grids are built, not bought,” stresses Murphy. “When we talk about grid computing, we're really talking about laying the groundwork for the evolution of your current enterprise-wide distributed computing environment, not a magic black box that you order and deploy.”

Keys to understanding the promise of grids

Key to understanding the promise of grid computing, she adds, is the realization that a grid — as it relates to IT — is a software environment based on open standards and protocols that make it possible to share disparate, loosely coupled IT resources across organizations and geographies. In a grid, resources such as computer cycles, storage capacity, databases, applications, files, sensors or scientific instruments, can be dynamically provisioned to the users or applications that need them.

One reason for confusion in the marketplace, is that different organizations are at varying stages of the grid evolution, depending on their own set of unique circumstances, adds Niraj Srivastava, a senior technologist for the Advanced Solutions and Development Practice of HP Services. For example, they might begin with clusters and load balancing, or they could focus on administrative portals, data services and security issues.

No matter where you begin, pulling together components and protocols to build a grid is much like designing a house, says Srivastava. “You don't just order a house; you work with an architect to design the house that best suits your lifestyle and budget.”

For some organizations, the driving factor is the need to deploy a new IT infrastructure in support of a new business line. Others are looking at grids as a way to gain computational power they wouldn't be able to otherwise acquire. And still others are building grids to enable geographically dispersed users to collaborate. Determining those requirements up front should be the first step in any grid exercise.

“The components and protocols that make up a grid aren't always new technologies,” explains Srivastava, referring to resource managers, portals, schedulers and clusters. “Building a grid infrastructure is about pulling them together in a modular framework based on standards.”

Benefits of grid computing

From an IT perspective, one of the key benefits to deploying a grid infrastructure is the ability to pinpoint under-utilized resources. In large pharmaceutical companies, for example, grid technology is successfully being used to allow researchers located in different countries to share computational resources. As Srivastava explains, instead of each research lab building its own cluster of computational resources and then watching it sit idle for weeks or months at a time, they can deploy a grid strategy to share resources cross boundaries on an as-needed basis and with enhanced security. IT is able to get much more efficiency out of its infrastructure and researchers are able to bring drugs to market faster.

Other examples of grid computing are found in the scientific and technical research arena, including projects like CERN's Large Hadron Collider Computing Grid, used by a worldwide community of scientists; the Southern Partnership for Advanced Computational Infrastructures (SPACI), the first grid-enabled research and development IT environment in Europe; and, Canada's Shared Hierarchical Academic Research Network (SHARCNET), a computing grid serving 11 academic institutions.

HP and grid computing

As a grid pioneer and major contributor to these and other grid projects, HP is committed to bringing the benefits of grid computing to the enterprise. HP has ported, tested and tuned the Globus Toolkit — the open source grid infrastructure implementation — to HP server platforms, and works closely with independent software vendors such as United Devices, DataSynapse, 

Platform Computing, Altair Engineering, NICE and TurboWorx to ensure third-party grid software works well on HP platforms. It also provides the management software and storage architectures that enable flexible, scalable grid solutions.

In addition, HP Consulting and Integration Services provides a comprehensive range of services to help plan, design and implement a grid computing infrastructure, including:

  • Grid Orientation
  • Grid Assessment Workshop
  • Grid Knowledge Transfer Workshop
  • Grid Design and Deployment Services
  • Grid Optimization Service

“Grids are no longer a vision for the future,” says HP's Murphy. “They're here today, but it's still early on in the evolution.”

More information on HP grid technologies and services

Download an HP grid white paper and solutions brief. See: HP Grid Technologies and Services

View an HP webcast discussing the problems solved by grid, including example use cases
Printable version
Privacy statement Using this site means you accept its terms
© 2008 Hewlett-Packard Development Company, L.P.