Why settle for a normal desktop computer when you can have a supercomputer instead? And I’m not talking about those gigantic machines. I’m referring to the latest development in the University of Maryland’s A. James Clark School of Engineering.
There, a group of researchers led by Uzi Vishkin have managed to develop a desktop computer that has a normal-sized circuit board, but is 100 times faster than average desktops. By using parallel processing on a single chip, these pioneers have managed to build a prototype model of their next generation personal computer. That their feat is remarkable can be gauged from the fact that they’ve mounted 64 parallel processors on a circuit board of the size of a license plate.
So, what exactly is parallel processing?
Parallel processing, as distinguished from the conventional desktops’ serial processing, is a technique that allows computer to perform a variety of tasks concurrently. It is but obvious that this results in a higher computing speed.
Parallel processing bettered in Desktop Parallel processing
Parallel processing has been used on an immense scale from years by using several interconnected chips or computers to develop supercomputers. However, its application to desktop systems was a challenge, due to severe programming difficulties. To combat those complexities Clark School team has devised a new method by using a single chip parallel processing technology.
Vishkin, a professor in the Clark School’s electrical and computer engineering department and the University of Maryland Institute for Advanced Computer Studies (UMIACS), explained the advantage of parallel processing by asserting:
Suppose you hire one person to clean your home, and it takes five hours, or 300 minutes, for the person to perform each task, one after the other. That’s analogous to the current serial processing method. Now imagine that you have 100 cleaning people who can work on your home at the same time! That is the parallel processing method. The ’software’ challenge is: Can you manage all the different tasks and workers so that the job is completed in 3 minutes instead of 300.Our algorithms make that feasible for general-purpose computing tasks for the first time.
That puts everything in perspective, doesn’t it?
Now, Vishkin and his team are looking ahead to showcasing their technology in future devices that will incorporate 1,000 processors on a fingernail-sized chip. To show how cool, this new program is, the team is also allowing students at Montgomery Blair High School in Montgomery County, MD to access and analyze their prototype.
Fiction to truth
Though, personal computer industry achieved success in computer clock speed via advancements in chip fabrication technologies and miniaturization. As per Moore’s Law, the numbers of transistors in integrated circuits of computers will double every 18 to 24 months with a parallel improvement in clock speed.
Vishkin, also known as inventor of parallel computing, commenced his work in 1979 by developing a theory of parallel algorithms based on a mathematical model of a parallel computer when no practical parallel prototype was there. 1997 onwards, he had tried to build a prototype to test his theory.
The hardware attributes of the sample device built are remarkably ordinary. What is remarkable is device’s parallel architecture, simplicity of programming and processing performance compared to other computers with the same clock speed.
Vishkin further asserted:
Based on the very positive reactions of my graduate students this spring, I knew that it was time to take the technology public.
Vishkin, in conjunction with his Ph.D. student, Xingzhi Wen, also published a hand out earlier this month about his newly built parallel processing technology for the Association of Computing Machinery (ACM) Symposium on Parallelism in Algorithms and Architectures. He also exhibited this innovation at a major computing conference - ACM International Conference on Supercomputing (ICS) in Seattle. In the conference he offered colleagues and student participants a chance to play with the prototype. He facilitated participants to bond the device remotely to run programs on it in a session lasting one whole day.
He also participated in a panel discussion with a special invitation sent from Microsoft Workshop on Many-Core Computing on June 20-21 in Seattle,Wash. And in August Vishkin will present an important speech at the Workshop on Highly Parallel Processing on a Chip in Rennes, France that will take place in collaboration with the 13th Euro-Par, an international European conference on parallel and distributed computing.
Since 1997 Vishkin has filed numerous patents on his parallel processing technology funded by the National Science Foundation and the Department of Defense. His ideas and research have received noteworthy interests from the computer industry.
Hunt for name:
To raise awareness of about this technology, Vishkin is also inviting public to propose names for it but on one condition that the name should compliment the features and bold aspirations of its parallel computing capabilities. The winner will also get $500 cash and a credit for doing so.
The deadline for submissions is September 15, 2007 and interested people can submit their ideas online.
Charles E. Leiserson, professor of computer science and engineering at MIT, compliments Prof. Uzi Vishkin by stating, ‘ Vishkin’s chip unites the theory of yesterday with the reality of today.” I’m inclined to agree, aren’t you?
Via: Physorg
