| Uncle Al 2005-02-02, 5:45 pm |
| Does anybody have slack time in a cluster they can volunteer? The
program runs as one parallel or multiple serial processes and is a
proven locator of defective nodes. Every CPU-week helps.
I have calculations of the handedness (normalized geometric parity
divergence) of a crystal lattice as its radius increases. We're
talking 10^15 atoms and more. Sparse radius sampling
http://www.mazepath.com/uncleal/combote.png
1.23x10^17 atoms for tellurium, 16 Opteron-848s, 10 days
is not sufficient. Dense radius sampling
http://www.mazepath.com/uncleal/qzdense.png
4.44x10^17 atoms for quartz, 150,000 CPU-hrs, completed
http://www.mazepath.com/uncleal/pddense.png
PdSbTe in progress
is CPU-intensive. This is a one-time shot for the last crystal
lattice. It supports an experiment to challenge a founding postulate
of General Relativity, qz.pdf below. Additional points on the PdSbTe
graph will require 10-20 Xeon-minutes each. You can appreciate how
adding 10,000 points is a right proper pisser.
Output looks like
Radius(A) Atoms CHI (parity divergence)
34988.000 7709943710067 0.999999995348076854
34992.000 7712588373872 0.999999994978445610
34996.000 7715233600906 0.999999997643456052
35000.000 7717879445838 0.999999999270916912
I will supply documented C++ source code to be locally compiled and
benchmarked. You want to know what is going on in there, right? This
is private inquiry - no funding, no commercial value, no security
issues.
Please post or e-mail me at the address in qz.pdf
Thank you!
--
Uncle Al
http://www.mazepath.com/uncleal/
(Toxic URL! Unsafe for children and most mammals)
http://www.mazepath.com/uncleal/qz.pdf
|