[MLUG] Even bigger days ahead for data superhighway

[Mike Miller <EMAIL:PROTECTED>]

What do you think of this?  Will they be able to keep this secure and
prevent it from being used for bad purposes?  If anyone can use it, for
any reason, won't it get clogged up with a bunch of garbage?  What about
worms -- won't they clog it up?

Mike


http://www.chicagotribune.com/news/local/chicago/chi-0309020132sep02,1,762170.story

Chicago Tribune
September 2, 2003

Even bigger days ahead for data superhighway

By Ronald Kotulak, Tribune science reporter

Imagine a Google-type search engine you could use at home to figure the
best odds on the Kentucky Derby or to predict weather in your
neighborhood.

The high-energy physics community that just 13 years ago made much of the
world's accumulated information available on the World Wide Web is now
working to make these problem-solving dreams come true.

Scientists at Fermi and Argonne National Laboratories, along with
colleagues around the globe, are building the ultimate supercomputer that
they envision almost anyone eventually could use to tackle tough problems
now out of their reach.

It is a system called the international data grid, which would link
computers in a new way to harness most of the world's number-crunching
capacity, along with tons of data now off limits.

The World Wide Web was invented out of the desperation of physicists who
were being overwhelmed by data that they couldn't share. The Web gave them
access to tons of data and revolutionized information transfer for the
rest of us.

Now physicists need the data grid and the computing power of thousands of
linked computers to grind the mountains of data down into nuggets of new
discoveries, an amazing capability that may eventually be more bountiful
to the average computer user than www.

It is also expected to break the growing stranglehold on scientific
breakthroughs now held by countries that can afford to build massive
particle accelerators, orbiting telescopes and scientific combines to
decipher the human genome.


Choked by data

"It's a democratization of science," said Fermilab scientist Lothar
Bauerdick. "The resources will be there and you can tap into them. It
gives equal opportunities to everyone to become part of science and part
of the discoveries. It started as a need in science but offers a
transforming technology for society."

Even rich countries are being choked to a standstill by the enormous
amounts of data generated in the fields of physics, astronomy, biology,
meteorology and Earth observations. Mounds of valuable data are being
built like expensive race cars that have few roads to travel on. The grid
will provide a network of interconnected highways.

"The data grid will call more brain power into science," said Vicky White,
chief of Fermilab's computer division. "What happens now is that we're
behind in analyzing data. We get piles of data and it might take 2 years
to analyze it, and people have to come to a central lab to do it.

"We believe the data grid will enable us to do science in a way that will
be so much better than we've been able to do in the past," she said.

The spectacular growth of the World Wide Web was totally unexpected. The
Internet has grown 100 percent annually since 1988 and in the last few
years its use shot up 400 percent annually.

While computers started off processing kilobytes of information (equal to
about 1,000 bytes with one byte being 8 bits and a bit a 0 or 1) then
megabytes (about 1 million bytes) then gigabytes (about 1 billion bytes),
they are now being asked to handle terabytes (1 trillion bytes).

When the mighty CERN accelerator goes online in Geneva, Switzerland, in
about 5 years, it will churn out petabytes (about a quadrillion bytes) of
data. A 1-year run of CERN data will be equal in volume to all the books
in the world's academic libraries.

"The goal is to accelerate the handling of the dramatic increase in the
amount of data scientists have to deal with," said Ian Foster, a math and
computer science expert at Argonne and the University of Chicago.

"In the face of this data explosion the capability of an individual
scientist to actually ask and get answers to interesting questions has
become challenging," said Foster, considered the father of the data grid.
"The data grid will allow them to get answers by harnessing large numbers
of computing and storage resources to perform their computations."


First came the Internet

Physicists faced a similar roadblock in the late '80s. Individual
laboratories were compiling massive amounts of data, but it was often
useless because scientists at other labs couldn't access it.

All they had was the Internet, which took off in the mid-'80s, and e-mail.
Then in 1990 CERN scientist Tim Berners-Lee invented the Web, making it
possible to link documents, graphs and pictures to other computers.

On Dec. 12, 1991, Berners-Lee and Paul Kunz of the Stanford Linear
Accelerator Center clicked on the Web for the first time at a meeting of
scientists, showed how easy it was to search the database, and launched an
information transfer revolution that has become integral to all parts of
society.

The twin problems facing physicists today are lack of access to monumental
stockpiles of data and an easy way to compute the data.

Foster and Carl Kesselman of the University of Southern California,
co-authors of the pioneering book "The Grid: Blueprint for a New Computing
Infrastructure," compare the status of computation to electricity around
1910.

Electric power generation was available then, just as the Internet is
available today. New devices were being invented that depended on electric
power, but their use was limited because they required each user to build
and operate an electric generator.

The truly revolutionary development was not electricity, Foster said, but
the electric power grid and the associated transmission and distribution
technologies.

Setting up standards for power production and connecting power stations
over wide swaths of the country made electricity available on demand
anywhere, anytime.

Millions of computing and storage systems exist all over the world that
are connected through the Internet. But as in 1910, these systems are
generally operated independently and most do not share access to their
computing power or reams of data.


Keeping hackers out

Security is a major reason for this roadblock. But Foster and others say
they are developing ways to allow access, at first, only to authorized
users and to keep out hackers. As security measures improve, access to the
international grid is expected to become available to an increasing number
of users.

The idea behind the data grid is to make computing power and access to
data as simple to use as electricity. Linking thousands of computers
together at major centers around the world would make enormous computing
power available to even the remotest user.

"Behind the scenes this huge machinery would spring into action," White
said. "You'd sit there and go for coffee maybe and all over the world your
request would have gone through masses of data and compared stuff and come
back with the answer," White said.

Trial runs between Fermilab and other centers in the United States and
Europe show that the grid is feasible.

"The vision of the grid is that it would provide computational power on
tap," White said. "If someone wanted to run through a simulation for
whatever purpose, betting on horses, playing the stock market or doing
high-energy physics in Calcutta, they would get access to those services
instead of having to go down and buy their own supercomputer."


Copyright ) 2003, Chicago Tribune
_______________________________________________
members mailing list
EMAIL:PROTECTED
http://mlug.missouri.edu/mailman/listinfo/members