[time-nuts] How good is the left end of your ADEV curve?
jimlux at earthlink.net
Wed Jan 25 22:24:49 EST 2017
On 1/25/17 6:58 PM, Tom Van Baak wrote:
>> What can I do at home, to observe such processes? Or is it way beyond
>> any imagination to participate in any such experiments?
> LIGO is a billion dollar experiment, involving thousands of PhD's so it will be some time until you can do that sort of stuff alone at home, or with your family.
> Jim Palfreyman has mentioned before what it would take to do Pulsar measurements as a home experiment. Search for the old threads or he can jump in to remind us why it can't or hasn't been done yet. See also the thread a month ago about a DIY H-masers since you'll want some of them on hand before you start.
> It's worth spending time reading anything about LIGO. The experiment is out-of-this-world clever, complex, sensitive. And it actually works! Unlike the particle physics tree, which seems to be nearing the end of bearing fruit, LIGO is at the very beginning of an entirely new way to study the universe.
I wonder if there are ways to do this kind of science in a massively
parallel way.. rather than the "one big awesomely high performing
instrument" you have a million mediocre instruments...
Of course, I know that doesn't always work, otherwise we could just buy
1000 cheap crystals and tell the maser folks to peddle their wares
But, as in many other endeavors, there's a limit to "how big/fast/good"
a single device can be, and you have to go to multiple devices - there's
always complexity and a learning curve, but eventually there is success:
One big power grid tube is better than many smaller ones, but
eventually, you hit the maximum size tube, and if you need more power
there's nowhere else to go but multiples.
Scientific computation hit the "single processor" wall, ultimately
resulting in the development of modern Beowulf cluster computers, which
in turn forced the development of new algorithms and reformulating the
underlying problem to allow such large clusters to be useful (Amdahl's
law, and all), and now things like exascale computing are becoming reality.
I've thought about whether one could do amateur radio Venus bounce or
Mars bounce, with a distributed transmitter/receiver system, timed by
GPS, so that you can do coherent processing.
More information about the time-nuts