Ian Parker
Jan20-09, 06:00 AM
On 17 Jan, 16:55, Benjamin Gittins <b.gitt...@synaptic-labs.com>
wrote:
> Hi,
>
> I have written up a short article on my website which asks the
> question: How long can we wait before we must take steps to protect
> our selves/infrastructure from quantum computer attacks?
>
> http://synaptic-labs.com/ecosystem/context-qc-relevant-today.html
I think a number of points need to be made here.
1) The QM computer is indeed designed to factorize numbers. However
you will not be able to factorize the products of large primes just
like that. At present only relatively small numbers have been
factorized.
http://en.wikipedia.org/wiki/General_number_field_sieve
Represents the state of thec art. Is it the final state of art?
http://en.wikipedia.org/wiki/RSA_Factoring_Challenge
We see from the above reference that the prize is no longer under
offer RSA seems to believe that the GNFS cannot be improved upon.
2) The QM computer will come in gradually. If required the length of
key can be increased. There is one approach that could in the future
be used and that is to use a very long (32K bits or so) number to
start off communication with a generating function used for subsequent
communication. A generating function is perfectly secure, but you
cannot transmit it unencrypted. Both computers have the same
generating function and perform an exclusive or.
Of course the most intersting thing from the point of view of academic
viewpoint is whether in fact the time we would have to run the
computer for is related in any way to the length of integers.
7*3 = 21 This seems to be about the current level. We can look at this
another way. If we input "21" it causes the Hamiltonian of 7,3 to be
different from that of other numbers. A Hamiltonian is, of course, a
matrix and what we have done is construct of Hamiltonian with its
matrix being logical extressions.
The question I would ask, I am afraid I don't really have the
wherwithall to answer is this. "Is quantum mechanics an escape from
the normal constraints of complexity theory?" A Hamiltonian does of
course have topological characteristics. If the answer in "no" there
are no security implications.
- Ian Parker
wrote:
> Hi,
>
> I have written up a short article on my website which asks the
> question: How long can we wait before we must take steps to protect
> our selves/infrastructure from quantum computer attacks?
>
> http://synaptic-labs.com/ecosystem/context-qc-relevant-today.html
I think a number of points need to be made here.
1) The QM computer is indeed designed to factorize numbers. However
you will not be able to factorize the products of large primes just
like that. At present only relatively small numbers have been
factorized.
http://en.wikipedia.org/wiki/General_number_field_sieve
Represents the state of thec art. Is it the final state of art?
http://en.wikipedia.org/wiki/RSA_Factoring_Challenge
We see from the above reference that the prize is no longer under
offer RSA seems to believe that the GNFS cannot be improved upon.
2) The QM computer will come in gradually. If required the length of
key can be increased. There is one approach that could in the future
be used and that is to use a very long (32K bits or so) number to
start off communication with a generating function used for subsequent
communication. A generating function is perfectly secure, but you
cannot transmit it unencrypted. Both computers have the same
generating function and perform an exclusive or.
Of course the most intersting thing from the point of view of academic
viewpoint is whether in fact the time we would have to run the
computer for is related in any way to the length of integers.
7*3 = 21 This seems to be about the current level. We can look at this
another way. If we input "21" it causes the Hamiltonian of 7,3 to be
different from that of other numbers. A Hamiltonian is, of course, a
matrix and what we have done is construct of Hamiltonian with its
matrix being logical extressions.
The question I would ask, I am afraid I don't really have the
wherwithall to answer is this. "Is quantum mechanics an escape from
the normal constraints of complexity theory?" A Hamiltonian does of
course have topological characteristics. If the answer in "no" there
are no security implications.
- Ian Parker