|
|
|
Why can’t we do what Indian techies are doing
Where as Indian Prime Minister is becoming popular for dishing out “people-to people contact” mantra persistently followed by verbal brawls with the Pakistan leaders, contact between technologists, scientists and researchers from both sides have not yet materialized. If we browse at technological practice across the border we find lot of interesting work being done.
In the field of computers, an Indian engineer who leads a team of computer jockey at Virginia tech makes low-cost supercomputer. This has stunned the computing industry by jigging up one of the world’s fastest supercomputers in record time and at record low cost using off-the-shelf components.
It took Srinidhi Varadarajan (a graduate from a university in Hyderabad Deccan) and his crew at Virginia Tech only a month and $5million to configure a supercomputer from 1,100 Apple Macintosh machines. Typically, the fastest machines cost upward of $ 10 million—and up to $250 million—and they take months, even years, to build.
In benchmark tests last week, the Virginia Tech supercomputer, powered by 2,200 IBM microprocessors, could compute at 7.41 trillion floating point operations (teraflops) a second. That would place it fourth in the list of fastest supercomputers in the world, according to The New York Times, which first reported the story.
The machine had been powered up to 8.164 teraflops, pushing it up to a possible third place. And it’s only now operating at perhaps 50 per cent efficiency.
The record, by far, for the fastest machine is held by the Japanese Earth Simulator, a $250 -million behemoth that computes at around 35 teraflops. It is followed by two other supercomputers at the Los Alamos National Laboratory and the Lawrence Livermore National Laboratory, both dedicated to weapons design, and running at between 7 and 9 teraflops, according to a widely-accepted ratings list maintained by Jack Dongarra, a University of Tennessee computer scientist.
Mr Dongarra’s list, which is considered the A C Nielsen of the supercomputing industry, rates India’s home-made Param Padma supercomputer at 171st place in the Top 500. Only about half a dozen countries in the world have the ability to make, and have use for supercomputers.
However, Mr Dongarra’s list is constantly being challenged and updated as even supercomputing becomes faster and cheaper all the time, broadly following Moore’s Law, which postulated that the number of transistors in a chip would double every eighteen months or so.
According to Mr Varadajan, supercomputing may now be improving even quicker, perhaps at the rate of 35 per cent every six months. Already, the Virginia Tech team is keeping a keen eye on a machine at US Department of Energy’s Pacific Northwest National Laboratory which claims to have run at 11.8 teraflops.
While three or four fastest supercomputers in the US all belong to Department of Energy labs, Virginia Tech’s breakthrough in the field occurred without planning, says Mr Varadarajan.
The school had a 200-node system which was becoming too small for its computing needs, when it wrote for a grant to build a new machine. When the proposal was being finalized, Apple came out with its new 64-bit G-5 Mac which the team found to its liking.
“Once we decided to build the supercomputer, we wanted to cut down the build cycle time of two years,” Mr Varadarajan explained. “What’s the point in buying at today’s prices and getting it to work many months later when the prices would have come down even further?”
The result was one month of almost non-stop work during which team members, including many students, were given free football tickets and endless supply of free pizza. The result is a $5 million machine that looks set to galvanize the industry.
Both the Los Alamos and Livermore supercomputers are said to have cost more than $15 million each.
The breakthrough in time and cost is also expected to revitalize the supercomputing industry that went through a downswing in the 1990s, including the near demise of companies such as Silicon Graphics, which makes the Cray supercomputer, two of which India bought in the late 1980s.
But the big irons are now making a comeback with computational science being used to simulate the behavior of natural or human engineered system, rather than just to observe the system or build a physical model of it like in the old days. Virginia Tech itself hopes to use its new computing muscle in areas such as nanoscale electronics, quantum chemistry, computational acoustics, and the molecular modeling of proteins.
The Virginia Tech machine also uses a unique cooling system and a fault tolerant software programme, called Déjà vu.
In the areas of Nanotechnology, US giant General Elelectric — popularly known as GE — has set up facilities in India, .and is going for technology called Cool Technology using Indian techies. GE past chairman Jack Welch was voted as most successful leader in the US business world partly because of his vision, was responsible for technical investment in India.
While other companies are thinking bigger is better, GE Global Research is thinking smaller. At least, that’s what the folks say about nanotechnology program. Realizing the enormous impact of breakthroughs like atomic-scale circuits, nanotube computers and self-assembled block copolymers, long-term investments in these technologies are being made to change the world - little by little.
Core technologies include:
— Nanotubes and nanowires
— Nanocomposites
—Nano-structured optoelectronics
— Biomimetics
“Every day we push technology to its limits — sometimes inch by inch and occasionally in great leaps forward. Our job is to make the impossible seem possible,”say GE.
Each improvement, every breakthrough, carries vast implications. A quieter, more powerful jet engine powers the planes we fly in. An energy-efficient gas turbine lights our cities and towns. An all-digital medical imaging system helps doctors save lives. These are the types of advances GE Global Research brings to life. This is what is called as cool technology.
Wearing helmet is now a law both in India and Pakistan as it makes the difference between life and death India with its vast population needs more helmets then any other country.
A new process for manufacturing helmet called the BPA (bisphenol A) has been developed, the chemical intermediate that is at the heart of the impact-resistant Lexan. In addition to helmets and impact-resistant eyeglass lenses, BPA is the fundamental building block in the manufacture of epoxy resins and polycarbonate plastic products such as food and beverage containers, compact discs, auto-glazing, and other consumer products.
At GE Global Research, techies are going one step further by looking for ways to manufacture BPA at a lower cost while maintaining the same high quality expected by our customers. A 25-person global team, including 16 scientists from the John F Welch Technology Centre in Bangalore, India, began an 18-month project in January 2001 to develop a new BPA process. The goal? To take $50 million in cost out of fixed investments in the future BPA facility while developing a simpler and more environmentally-friendly process.
The new process, which is being developed from the ground up, also will provide an opportunity to improve the efficiency of the current global BPA manufacture via a relatively simple retrofit, as well as showcase innovation and direct R&D impact on a business in the best possible light. At GE Global Research in Bangalore it is said that the future is so bright that you’ve got to wear shades — made from BPA.
In areas of materials beyond super alloys exploiting high-temperature composites research is being done on temperatures of blade surfaces in advanced aircraft engines which are approaching super alloy limits. Innovations in refractory metal-intermetallic composites (RMIC’s) are being pursued in systems based on Nb-Ti-Si and Mo-Si-B.
These systems have the potential for service at surface temperatures greater than 1350oC. If either system is to be employed, a balance of properties is required for many different behaviors that must be integrated into a workable design, and affordable fabrication of the composite components must be possible. Progress in the development of tough high-temperature composites for aircraft engine applications will be compared to performance needed to make these materials competitive choices for future turbine designs.
In space research Indian organization called Ispro, which is similar to our Suparco, is actively supporting research in the following categories:
Space science: Physics of the ionosphere and magnetosphere; meteorology, dynamics of the atmosphere, geophysics, geology, astronomy, cosmology, astrophysics, planetary and interplanetary space physics and climatology.
Space technology: Rocket and satellite technology; propulsion systems design and optimization; aerodynamics and heat transfer problems related to space vehicles; guidance and control systems for launch vehicles and spacecraft, polymer chemistry, propellant technology, ultra-light-weight structure, satellite energy systems, space electronics, space communication systems, orbital mechanics and computer sciences.
Space application: Remote sensing of the Earth’s resources — space communication, satellite geodesy image processing, satellite meteorology including weather forecasting.
We, in Pakistan, can benchmark and develop what Indian techies are doing. While we are having American investment in the form of KFC, Pizza Hut, and McDonalds, we still have to bring American technological firms such as GE and Dupont here. Can our worthy federal minister Shaukat Aziz do it? This is indeed a big question mark.
Now that we have gone nuclear, there is a strong need to look towards space technology. Though we may not have the support of, for instance, GE Global Research but we can tap Comstech funding. President of Pakistan is heading the Committee of Science & Technology of Islamic world. Can he use this platform for building space technology for the Muslim world? Just do it Pres!
The writer is a freelance contributor
|
|
|
|
