Turning Grid Computing Into A Viable Business Strategy

Updated: Nov 11 2003, 05:30am hrs
In the wake of the rise and fall of the dot-com era, new technologies that promise businesses the moon and the stars are viewed with a healthy dose of skepticism. There has been a lot of hype recently around grid computing, leaving IT managers wondering if it really is a game-changing approach to computing, or just another Next Big Thing.

Like the Internet, grid computing grew out of research and academia and has moved into commercial markets. But before it can take hold as a viable business computing strategy, we need to be clear on what grid computing is and the value it brings to commercial enterprises. People typically think of a computing grid as a kind of virtual supercomputer clusters of geographically distributed systems linked together and focused on a single, compute-intensive task.

An example often cited from the world of science and research is SETI@home, which uses excess processing cycles on PCs to examine radio signals for signs of life on other planets. The idea of a grid as just a supercomputer is light-years off course. Its true that many organisations are using grid technologies in a similar fashion to build virtual supercomputers that help them solve massive computational problems. This is one way but by no means the only way businesses can benefit from grid technologies. Most enterprises are focused on getting more value out of the computing and data assets they own. And thats the real value grid computing brings to the table.

The majority of the systems in enterprises are distributed midrange and desktop systems based on different architectures from multiple vendors and all too often, those systems dont work and play well together. Imagine if a business could connect those heterogeneous, widely dispersed computing and data assets, harness their collective processing power and manage them like a single large computer. Grid computing makes this possible.

Companies have made huge investments in computing capacity, much of which sits idle 80 to 90 per cent of the time. When those assets are grid-enabled, previously untapped capacity becomes available on an on-demand basis, similar to the way processing power within a single mainframe can be directed on-the-fly to where its needed. Many forward-thinking businesses are doing this today.

Financial services company Charles Schwab grid-enabled an existing financial application, reducing processing time on the application from more than four minutes to just 15 seconds. Based on the success of this project, the company is looking to expand grid research into other areas of its business. Similarly, grid pioneer Butterfly.net introduced a first-of-its-kind computing grid that enables online video game providers to reliably deliver state-of-the-art games to millions of concurrent users.

When the grid determines there are too many players connected to a server, it automatically reconfigures underutilised servers to support the most popular game-play and seamlessly transfers players to those servers. At its core, grid computing is a business thought. Just as the Web transformed the way we exchange information, grid computing takes distributed computing to the next level, giving businesses a unified, resilient and fully utilised IT infrastructure, available on demand.

Savvy companies are finding that grid computing allows them to manage workloads seamlessly across distributed systems and networks, giving them unprecedented access to their computing and data resources and the ability to manage those resources in ways never before possible. Grid computing is here today. Businesses that begin the journey to capitalise on its potential to drive productivity and efficiency up and operating costs down will have a significant competitive advantage.

The author is country manager for zSeries systems of IBM India