IBM Grid Computing











PDF version of this report
You must have Adobe Acrobat reader to view, save, or print PDF files. The
reader is available for free
download
.

IBM
Grid Computing

by Sherry Kercher

Docid: 00018707

Publication Date: 2006

Report Type: PRODUCT

Preview

IBM’s grid computing software and middleware are designed for intensive
data processing. Grid computing combines the capabilities of multiple PCs
and servers to perform complex computations. IBM has begun promoting
“smart grids” for utilities to increase awareness of energy usage. IBM
also sponsors the World Community Grid, which pools PC resources from
global volunteers to solve problems related to disease, food distribution,
and other concerns. IBM is currently hosting a Scripps Research project to
help scientists virtually screen chemical compounds to help battle
COVID-19.

Report Contents:

Description

[return to top of this
report]

IBM offers grid computing software, middleware, and services to power
distributed computing systems for enterprises, researchers, and nonprofit
organizations.

 


Related Faulkner Reports
IBM Company
Profile

Grid computing has emerged as a way to harness and take advantage of
computing resources across locations and organizations. Grid computing
refers to the use of numerous PCs and servers to deliver higher levels of
computing than would be possible on just one. In this model, computing
resources are pooled across an enterprise, organization, or even global
community. Hypothetically, grid computing offers limitless possibilities
for computing power, making it ideal for research, analysis, and
large-scale processing.


Vendor

Name: IBM
Headquarters
1 New Orchard Road
Armonk, NY 10504
(914) 499-1900
(800) 426-4968
Web: http://www.ibm.com/
Type of Vendor: Computer and Communications
Hardware, Software, and Services
Founded: 1914
Service Areas: Worldwide
Stock Symbol: IBM (NYSE)

Grid computing can be used to consolidate a company’s workload
management, provide capacity for high-demand applications, and reduce
cycle times. It can also increase data and collaboration access by
federating data, supporting large, multi-disciplinary collaboration, and
enabling inter-organizational and business collaboration, also making a
company’s infrastructure highly available by balancing workloads and
accounting for recovery and failover.

Platform Symphony

IBM provides its Platform Symphony software to manage the performance of
distributed applications and big data analytics on a shared grid. The
software is designed to reallocate over 1,000 compute engines per second
to different workloads depending on pre-defined policies.

Platform Symphony Advanced Edition offers an Apache Hadoop-compatible
MapReduce implementation to deliver approximately four times the
performance of open source Hadoop. The Platform Application Service
Controller for Platform Symphony add-on supports cloud native services
including Apache Spark, MongoDB, and Cassandra. Platform Symphony supports
multi-tenancy, which allows different workloads to dynamically share
resources according to policies. It can run on-premise, in the cloud via
IBM High Performance Services, or in a hybrid deployment.

In particular, IBM defines grid computing as the ability – using a set of
open standards and protocols – to access applications and data, processing
power, storage capacity, and other computing resources over the Internet.
A grid is a type of parallel and distributed system that enables the
sharing, selection, and aggregation of resources distributed across
multiple administrative domains based on resource availability, capacity,
performance, cost, and user quality of service requirements.

The IBM-sponsored World Community Grid (WCG) pools PC resources from the
idle capacity of systems owned by volunteers around the world to create a
giant virtual supercomputer. It has traditionally been used to research
diseases such as cancer or HIV/AIDS. IBM is also promoting smart grid
systems, which offers to help utilities or utility-style service providers
increase awareness of their energy usage and flow, to increase efficiency.

In general, IBM’s Grid Computing products can power demanding processes
such as DNA research, intensive mathematical computations, actuarial
analysis, medical records storage, and company-wide data analytics. IBM’s
grid products include software and hardware implementations. Table 1 lists
some of IBM’s top Grid Computing product offerings.

Table 1. Grid Computing Offerings
Product Description Details
Grid Medical Archive Scalable virtualized storage for
healthcare, research, and pharmaceutical clients. It is designed
to provide enterprise wide medical image, data, and records access
while delivering real time business continuity.
Combines IBM TotalStorage and System x
servers with IBM System Storage Multilevel Grid Access Manager
software, on an open source operating system. Grid Medical Archive
can be used to:

  • Provide on demand access to medical images.
  • Eliminate data silos.
  • Foster business continuity.
  • Address regulatory disaster recovery considerations.
  • Support infrastructure resiliency with storage
    virtualization tools.
Grid and Grow Express Scalable, low end offering for small
businesses. It includes IBM and business partner technology along
with a “get started” services package to help customers implement
a grid computing infrastructure. 
Combines hardware, software, and
services for deploying a grid environment. Grid scheduler options
include Altair’s PBS Professional, Platform Computing’s LSF, or
IBM’s LoadLeveler.
Optimized Analytic Infrastructure Offers high velocity trading, risk
management, and sustainable performance. It was designed to
provide financial services firms with business and technical data
analysis, using IBM’s grid and high performance computing, and is
designed to manage applications, workloads, data, and systems.

Uses technology enhancements – such as System Optimizers – to
hone the platform’s capabilities for specific workload use. IBM
also provides pre-integrated technology that packages hardware
and software to support a specific workload or class of
workloads. Core elements include:

  • ApplicationWeb
  • General Parallel File System
  • Workload software
Smart Grid IBM’s initiative to apply grid computing
to the power industry. The system uses input and data collected
from “smart meters,” customer’s systems, and utilities’ systems to
balance and control power consumption, reducing the strain on the
connected power infrastructure, and reducing the amount of energy
being lost as waste. The technologies involved also figure
strongly in potential renewable energy initiatives, many of which
suffer from inconsistent generation of power. 
Smart Grid currently includes 4
separate initiatives and segments:

  • Customer Transformation Optimization – a
    service which provides detailed energy usage stats to
    customers
  • Intelligent Utility Network Solution – a
    product designed to improve efficiency, integrate renewable
    resources, and address climate change
  • Power Generation Optimization – a solution
    designed to improve plant operations, manage assets, and
    handle information
  • Solution Architecture for Energy and Utilities
    Framework (SAFE)
    – a planning service for
    integrating, managing, and optimizing utility systems. 
Grid Computing as Part of IBM’s On
Demand Strategy
IBM’s grid computing offerings are part
of an On Demand strategy for businesses, allowing them to
integrate, manage, and access business and IT processes from
anywhere across the company, as well as with partners, suppliers,
and customers. IBM’s On Demand strategy is designed to foster
responsiveness to business changes, market opportunities, and
variable cost structures, while supporting increased focus and
resiliency.
  • Standards Based Software – IBM Lotus
    collaboration software, database and software integration
    products, and IBM hosting and outsourcing services.
  • Grid Computing Services – Grid consultants
    and services for IBM grid users.
  • SOA Architecture Support – By design, IBM’s
    grid computing model supports a service oriented architecture
    (SOA) enterprise strategy of applications reuse. It supports
    common formats and database languages in WebSphere databases,
    the migration of COBOL to Java, and sharing of business logic
    across both the OLTP and Batch paradigms.

 

Table 2 outlines some of the grid computing projects and research
activities in which IBM is involved.

Table 2. IBM Grid Computing Projects
Project Description
World Community Grid The World Community Grid is a grid of PCs from
volunteers to form a virtual supercomputer, which provides
computing to solve the world’s complex community challenges.
Launched in 2004 by IBM, the WCG has connected more than 650,000
members and more than 2.7 million devices. IBM is hosting a
Scripps Research project to help scientists virtually screen
chemical compounds to help battle COVID-19.
The Globus Project IBM is a key collaborator in the Globus Project, the
multi-institutional research-and-development effort for grid. The
Globus Project is developing Open Grid Services Architecture
(OGSA), a set of standards and specifications that integrate Web
services with grid computing. IBM’s intraGrid, based on Globus, is
a research and development grid that allows IBM to integrate
multiple worldwide assets for research purposes, helping
understand the complexities of managing a grid infrastructure on
an enterprise scale.
Open Grid Forum IBM is a sponsor of the Open Grid Forum, the
mission for which is to develop industry standards for grid
computing.

Figure 1 is an infographic displaying statistics that IBM has made
available about the World Community Grid.

Figure 1. IBM World Community Grid Statistics

Figure 1. IBM World Community Grid Statistics

Source: IBM

Applications

[return to top of this
report]

IBM’s Grid Computing products and services were created to benefit
organizations looking for high power, distributed computing. These
organizations can include large businesses, researchers, and nonprofit
groups such as medical companies engaged in research and recordkeeping;
insurance firms doing risk analysis; and enterprises for general resource
planning. Table 3 outlines the main applications in IBM’s Grid Computing
portfolio.

Table 3. IBM Grid Computing Applications
Product Applications
IBM Grid Medical Archive

Healthcare, research, and pharmaceutical clients. 

IBM Grid Solution for Data Intensive
Computing

Large enterprises for initiatives such as CRM, compliance,
business intelligence, supply chain management, and overall
enterprise resource planning.

IBM Optimized Analysis Infrastructure Trading and financial services firms.
IBM Grid and Grow Express

Small to medium sized businesses.

Environment

[return to top of this
report]

Table 4 looks at hardware and software considerations for implementing an
IBM Grid product.

Table 4. Hardware and Software Support
Hardware and Network Considerations Software Considerations
IBM’s Grid Computing products operate
in an environment that can, hypothetically, connect an unlimited
amount of PCs and servers. IBM offers its own specific servers for
grid computing, which can be connected at a single location or
across multiple locations, using a secure network connection.
IBM’s software for grid computing is
largely based on Linux, and it is designed to operate with third
party software and hardware. Generally, IBM’s grid computing
environments can use most major operating systems, including IBM’s
as well as third party systems such as Microsoft Windows, Apple
Mac OS X, HP-UX, and assorted Linux platforms. Some of the
company’s sponsored grid computing projects also run on
third-party software, such as the World Community Grid’s BOINC
(Berkeley Open Infrastructure for Network Computing) operating
system.

IBM offers several support options for licensed customers. These options
include phone, fax, and Internet resources, as well as a Technical Library
for products, which contains data files, support guides, FAQs, product
downloads, and technical papers. Discussion forums and newsgroups are also
offered.

Pricing

[return to top of this
report]

IBM delivers its grid computing offerings both directly and indirectly –
either through its direct sales force or its international resellers and
qualified partners. These channels provide pricing information on request.

Competitors

[return to top of this
report]

IBM’s top grid competition comes from Cisco’s Smart Grid Solutions, HP
Grid Computing Solutions, and Oracle Grid Computing. These vendors have
delivered grid computing for many years, and continue to increase their
focus in the areas of smart grid utilities, medicine, and distributed
enterprise computing.

In addition, other Internet-based offerings, branded as cloud services,
have grown significantly in popularity in recent years and offer many
similar features and benefits. Competitors in this segment include grid
computing providers like HP along with many specialty competitors. IBM
offers its own cloud services, as well.

[return to top of this
report]

About the Author

[return to top of this
report]

Sherry Kercher is an editor for Faulkner Information
Services. She holds a master’s degree in library and information science,
and tracks and writes about storage, communication networks and equipment,
and Internet technologies.

[return to top of this
report]