Howdy Richard,
Digital computers generate pseudo-random numbers, which have limited
use. The info below is about *actual* random number generation.
Pseudo-random number generation is bad for use in gambling devices,
like video poker, because it gives the house unfair odds and
predictable games/patterns. It can give players an advantage if they
crack the number generator, because they can tell where it is in its
sequence by observing game play. Ordinary pseudo-random number
generators are also very bad for encryption mask generation, because
they are comparatively easily cracked using known plaintext attacks.
Important simulations should be checked using real random number
generation to avoid unexpected systematically skewed results. This is
an important field I think. I also think 2 Gbps is unimpressive
using today's circuitry.
BTW, networked X-box or Playstation game platforms make for
fantastically cheap supercomputers. Here's an example:
http://www.labspaces.net/94312/
PS_s_Help_Astrophysicists_Solve_Black_Hole_Mystery
http://tinyurl.com/6sr8qy
Tightly coupled multiprocessing is a very old discipline that pre-
dates supercomputing by decades. I did internals work over 25 years
ago on a highly commercial operating system designed to run 16
tightly coupled CPUs.
Sounds like somebody is giving you some overblown hype on quad
processing. I hope it is not a stock broker. It is true a lot of
graphics horsepower (and fast monitors) will be required when
operating systems go 3D, but this is not a technological breakthrough
AFAIK, except maybe for price. I kind of wonder if the new Sony
Bravia XBR7 with 240 Hz refresh rate is possibly designed for 3D
game playing. That gives the full two channels of 120 Hz image
bandwidth required for fast motion 3D. The remaining hardware need
is a pair of synchronized glasses. I think there are some wild 3D
games and videos coming! The new supercomputer architecture game
platforms are very much up to the job for that, and already very cheap.
Best regards,
Horace Heffner
http://www.mtaonline.net/~hheffner/
On Dec 28, 2008, at 5:13 PM, R C Macaulay wrote:
Howdy Horace,
May work for a short time for incripting devices but quadratic
computing ( use of one main, one slave and two image tracking
computers makes child's play of nearly any incription strategy. An
example of a "quad" setup could be as simple as having two image
trackers deal with (7) to the 21st power each. If you want to take
it higher add a 3rd image tracker.
Richard
We've discussed random number generation schemes here before. They
are important to simulations, experimental design, gambling devices,
and cryptography. Here is one that works at 1.7 gigabits per second
(Gbps):
http://www.physorg.com/news148660964.html
The above scheme works based on sensing a potential, specifically a
photodetector potential, which is a process subject to hysteresis,
and thus bias, when converting two ranges of potentials to
corresponding bit values. As discussed before, there is a solution
available for this bias problem:
