Glenn writes:
> I was made aware this weekend of some tips for running the SETI@home
> client under MS Windows:
>
> (1) The SETI 2.0 command line client for NT works just fine on Win95
> and Win98. Scrap your screensaver if you really want to crunch
> and just run this bad boy in the background. Much faster than
> the GUI version.
Actually, I've tried both, the "command line client" and the GUI version --
and I haven't found any real speed difference between the two, once the GUI
version goes to a blank screen and is run in background full-time.
However, what you do miss if you run only the CLC version is the science in
the displayed fast fourier transforms (besides, it also takes some of the fun
out of the process and makes the entire enterprise into a "scientifc pissing
contest," as Goetz Neumann said quite a while ago :-).
If you've downloaded the newest version of the GUI client (version 2 or
better), they're now displaying the largest gaussian peak that your work unit
finds. These peaks represent the broad emission lines of neutral hydrogen in
the galactic disk. I'm actually sort of surprised that they chose this
frequency range, simply because it is so noisy.
The emanations you're seeing on your screen are the result of billions of
billions of hydrogen atoms engaging in a anti-parallel spin flip, which if
you remember your quantum mechanical chemistry well, you know to be the
smallest possible energy change in quantum numbers for a neutral hydrogen
atom, when at the ground state.
This hydrogen gas is the primary constituent of what is called the
interstellar medium. For more information about the ISM, there is a
particularly nice web page at:
http://www.bc.cc.ca.us/programs/sea/astronomy/ismnotes/ismglxya.htm
Look for the section in the webpage on 21-cm radiation (a radio frequency can
be specified either by its cycles per second or its wavelength; the speed of
light is almost exactly 3x10^10 centimeters per second. When you divide
3x10^10 cm/s by 1.420 GHz, you get 21 cm. Astronomers tend to prefer
wavelengths over frequency when they write their notations).
By measuring the doppler shifts of the ISM's neutral hydrogen's radiation, a
map can be drawn of the Milky Way. Indeed, this is the only way that we any
idea of the shape of our own galaxy -- that, and a presumption that the
Andromeda galaxy, which is right next door (so to speak), is a close relative.
However, the doppler calculations that give us some sense of the map are
greatly conflated by the local motion of the Earth. The antenna's position
against the sky and the time of day all have to be taken into account to get
a proper relative motion of the particular patch of hydrogen that is being
observed in any one work unit.
As I say, I've always been a little surprised that they chose this frequency
band to observe. It would be much harder to punch a signal through this noise
than it would be if a quieter part of the spectrum had been chosen (in the
movie, "Contact," the alien signal was detected at "hydrogen times pi," a
very reasonable point in frequency space, for several different reasons).
On the plus side, what may come out of all of this is nothing more than a
really good map of the Milky Way, and that may be all that the people who set
this program up were expecting.
> Team HP3000-L is holding at #77 in the "Clubs" category, being remarkably
> stable almost since breaking into the top 100 over six months ago.
On a second, more mathematical/physical note, Glenn's observation of the
"remarkable stability" of this contest is what should be expected. Indeed,
after a fashion, this entire counting enterprise replicates an expanding
universe, to at least a small degree.
Because everyone that is participating -- and because almost everyone that is
participating is male (meaning that they generally are competitive and adopt
a "take no prisoners" attitude) -- they generally reached their processor
saturation capabilities almost immediately after signing up. That condition
of saturation fairly well set in concrete from the very first day the
relative powers of the various contestants. With the exception of a few minor
exchanges here and there due to machines coming on- or off-line*, the
patterns that we started with are going to be very similar to the patterns we
finish with, with the only caveat being the total counts.
As a consequence, the process is exactly like riding on the surface of an
inflating balloon. The relative sort order of the various contestants won't
change much from now on -- only the distance between various entrants will.
Wirt Atmar
*for example, Mark Klein must have just added a few new machines to his crew
because he's now been inextricably gaining on me for the last week or so.
However, I would never stoop so low as to either notice that, much less
mention it.
|
