Seti: Is Anybody Out There?
by Leander Kahney
Wired News
Dec. 22, 2000
Seti@Home has clocked half-a-million years in computer time searching for
ET on home PCs, but the real work isn't due to begin in earnest until January.
The Seti@Home project harnesses the spare computing cycles of millions of
PCs across the globe to search for telltale signs of intelligent life in
radio signals beamed from outer space.
Since its launch 18 months ago, 2.6 million people in 226 countries have
downloaded Seti@Home's screensaver software, which scans radio data when
their computers aren't otherwise in use.
The screensaver looks for strong signals among the white noise of the
universe's background radio transmissions.
But the process of figuring out which signals, if any, have come from alien
civilizations won't start operating at full swing until the end of January,
when the project brings online all its back-end servers.
"We only recently started the back-end analysis," said David Anderson,
Seti@Home's director. "We've been doing bits and pieces here and there. It
will be turned on full blast next month."
In the meantime, the Seti@Home project has become the largest computing
problem ever undertaken. It has clocked an astonishing 500,000 years in
computing time, with another 1,000 years of processing time contributed
every day.
Collectively, the 2.6 million Seti@Home machines are twice as powerful as
the most powerful supercomputer on the planet, and a lot cheaper.
The U.S. Government's ASCI White, housed at a California nuclear weapons
research lab, is rated at 12 Teraflops (trillion operations a second) and
cost $110 million to build.
All day, every day, Seti@Home operates at around 25 Teraflops but cost only
$500,000, excluding the cost of the home PCs.
"It's the biggest supercomputer on the planet by far," said the project's
chief scientist, Dan Werthimer, an astronomer at the University of
California at Berkeley's Space Sciences Laboratory.
To date, the Seti@Home screensavers have identified 500 million strong
radio signals from data gathered by the Arecibo radio telescope in Puerto
Rico, the world's largest radio telescope, which was built into a huge,
bowl-shaped sink hole at the top of a mountain.
"In the screensaver, strong signals look like skyscrapers sticking up above
the grass," said Werthimer.
Using a variety of algorithms, the screensavers have whittled down the 500
million strong signals, or spikes, to about 125 million that look interesting.
All of these 125 million signals display a Gaussian, or bell curve,
profile. As the Aricebo telescope scans across the sky, the signal becomes
progressively stronger and then weaker, just like a bell-shaped curve.
A signal displaying a Gaussian profile is less likely to be noise from
terrestrial cell phones, airplanes or satellites, and probably originated
from a single point in space -- maybe a planet orbiting a nearby star.
"That characteristic pattern gets us excited," said Werthimer.
But to ensure the signal came from an alien civilization, it must be
transmitted from the same part of the sky, at the same frequency, on two or
more observations, which would indicate a persistent signal and not just a
random burst of energy.
This "back-end" analysis was only recently begun, but already it has
identified hundreds of signals that have been broadcast twice, three times,
even four times, over an 18-month period. A map on the project's site shows
the results: Red dots are signals picked up more twice. Green dots are
signals detected four or more times.
But so far, every signal that merits closer inspection has proven a red
herring.
"Occasionally you find very strong signals and you look more closely but it
turns out to be a satellite or interference," said Werthimer.
Even though 500,000 years of computing time seems like a lot of time to
throw at a problem, Werthimer said it's just the beginning.
"I've been doing this for 25 years and I still think we're just scratching
the surface," Werthimer said. "We've got a long way to go. We're searching
such a small range of frequencies, it may take another 50 or 100 years."
Seti@Home currently searches a narrow 2.5 Mhz band among the billions of
potential radio bands. A thorough search for signs of intelligent life
would include a much wider swath of the electromagnetic spectrum.
A good candidate would be light. Advanced alien civilizations might
communicate with lasers because of their high information bearing capacity.
Another Seti program at Berkeley plans to start searching for patterns in
light. But unlike radio, light has trouble penetrating the huge clouds of
cosmic dust that are spread throughout the universe.
There is also the problem of ever-increasing amounts of radio pollution.
The Earth is already brighter than the Sun in terms of radio radiation --
all of it man-made.
"It's getting harder to do Seti on this planet," said Werthimer. "We may
have to go to the back side of the moon."
The good news, Werthimer said, is that the ever-increasing power of
computers means that more and more cycles can be applied at the search.
Already Seti@Home has added the ability to find pulsed, and not just
steady, signals, and will shortly include data from the Southern Hemisphere.
Werthimer said Seti's search capabilities have increased by a factor of one
million in the last 20 years, but it will take a one billion-fold increase
before a comprehensive search of the universe can be done.
"We are 20 or 30 years away from a thorough search," he said. "It's like
combing a cosmic haystack. We've just started poking around the edges."
"Even though I think the universe is teeming with life, I think it's going
to be a while. But it will probably be in our lifetime."
Seti@Home's director, Anderson, added: "Detecting an alien signal would
probably be one of the most exciting events in human history, if not the
most exciting.
"It would change all of our lives forever. The probability of it happening
is not exceptionally high but it's like the lottery. The payoff is so big,
it's worth doing anyway."
by Leander Kahney
Wired News
Dec. 22, 2000
Seti@Home has clocked half-a-million years in computer time searching for
ET on home PCs, but the real work isn't due to begin in earnest until January.
The Seti@Home project harnesses the spare computing cycles of millions of
PCs across the globe to search for telltale signs of intelligent life in
radio signals beamed from outer space.
Since its launch 18 months ago, 2.6 million people in 226 countries have
downloaded Seti@Home's screensaver software, which scans radio data when
their computers aren't otherwise in use.
The screensaver looks for strong signals among the white noise of the
universe's background radio transmissions.
But the process of figuring out which signals, if any, have come from alien
civilizations won't start operating at full swing until the end of January,
when the project brings online all its back-end servers.
"We only recently started the back-end analysis," said David Anderson,
Seti@Home's director. "We've been doing bits and pieces here and there. It
will be turned on full blast next month."
In the meantime, the Seti@Home project has become the largest computing
problem ever undertaken. It has clocked an astonishing 500,000 years in
computing time, with another 1,000 years of processing time contributed
every day.
Collectively, the 2.6 million Seti@Home machines are twice as powerful as
the most powerful supercomputer on the planet, and a lot cheaper.
The U.S. Government's ASCI White, housed at a California nuclear weapons
research lab, is rated at 12 Teraflops (trillion operations a second) and
cost $110 million to build.
All day, every day, Seti@Home operates at around 25 Teraflops but cost only
$500,000, excluding the cost of the home PCs.
"It's the biggest supercomputer on the planet by far," said the project's
chief scientist, Dan Werthimer, an astronomer at the University of
California at Berkeley's Space Sciences Laboratory.
To date, the Seti@Home screensavers have identified 500 million strong
radio signals from data gathered by the Arecibo radio telescope in Puerto
Rico, the world's largest radio telescope, which was built into a huge,
bowl-shaped sink hole at the top of a mountain.
"In the screensaver, strong signals look like skyscrapers sticking up above
the grass," said Werthimer.
Using a variety of algorithms, the screensavers have whittled down the 500
million strong signals, or spikes, to about 125 million that look interesting.
All of these 125 million signals display a Gaussian, or bell curve,
profile. As the Aricebo telescope scans across the sky, the signal becomes
progressively stronger and then weaker, just like a bell-shaped curve.
A signal displaying a Gaussian profile is less likely to be noise from
terrestrial cell phones, airplanes or satellites, and probably originated
from a single point in space -- maybe a planet orbiting a nearby star.
"That characteristic pattern gets us excited," said Werthimer.
But to ensure the signal came from an alien civilization, it must be
transmitted from the same part of the sky, at the same frequency, on two or
more observations, which would indicate a persistent signal and not just a
random burst of energy.
This "back-end" analysis was only recently begun, but already it has
identified hundreds of signals that have been broadcast twice, three times,
even four times, over an 18-month period. A map on the project's site shows
the results: Red dots are signals picked up more twice. Green dots are
signals detected four or more times.
But so far, every signal that merits closer inspection has proven a red
herring.
"Occasionally you find very strong signals and you look more closely but it
turns out to be a satellite or interference," said Werthimer.
Even though 500,000 years of computing time seems like a lot of time to
throw at a problem, Werthimer said it's just the beginning.
"I've been doing this for 25 years and I still think we're just scratching
the surface," Werthimer said. "We've got a long way to go. We're searching
such a small range of frequencies, it may take another 50 or 100 years."
Seti@Home currently searches a narrow 2.5 Mhz band among the billions of
potential radio bands. A thorough search for signs of intelligent life
would include a much wider swath of the electromagnetic spectrum.
A good candidate would be light. Advanced alien civilizations might
communicate with lasers because of their high information bearing capacity.
Another Seti program at Berkeley plans to start searching for patterns in
light. But unlike radio, light has trouble penetrating the huge clouds of
cosmic dust that are spread throughout the universe.
There is also the problem of ever-increasing amounts of radio pollution.
The Earth is already brighter than the Sun in terms of radio radiation --
all of it man-made.
"It's getting harder to do Seti on this planet," said Werthimer. "We may
have to go to the back side of the moon."
The good news, Werthimer said, is that the ever-increasing power of
computers means that more and more cycles can be applied at the search.
Already Seti@Home has added the ability to find pulsed, and not just
steady, signals, and will shortly include data from the Southern Hemisphere.
Werthimer said Seti's search capabilities have increased by a factor of one
million in the last 20 years, but it will take a one billion-fold increase
before a comprehensive search of the universe can be done.
"We are 20 or 30 years away from a thorough search," he said. "It's like
combing a cosmic haystack. We've just started poking around the edges."
"Even though I think the universe is teeming with life, I think it's going
to be a while. But it will probably be in our lifetime."
Seti@Home's director, Anderson, added: "Detecting an alien signal would
probably be one of the most exciting events in human history, if not the
most exciting.
"It would change all of our lives forever. The probability of it happening
is not exceptionally high but it's like the lottery. The payoff is so big,
it's worth doing anyway."
