Author Topic: Quantum Progress  (Read 10301 times)

dedndave

  • Member
  • *****
  • Posts: 8828
  • Still using Abacus 2.0
    • DednDave
Quantum Progress
« on: October 10, 2012, 03:55:10 AM »
this is pretty cool....

http://www.euronews.com/newswires/1684922-french-us-scientists-win-nobel-prize-for-physics/

one downside that i see is that, if computers are super-fast, the coders will become super-lazy,
and we get super-crappy code

mineiro

  • Guest
Re: Quantum Progress
« Reply #1 on: October 10, 2012, 04:57:04 AM »
Nice link Sir dedndave;
I was thinking about how to program to that; maybe thinking in all possible situations? Only one possible path in the end, sound strange but sounds a nice challenge.

dedndave

  • Member
  • *****
  • Posts: 8828
  • Still using Abacus 2.0
    • DednDave
Re: Quantum Progress
« Reply #2 on: October 10, 2012, 05:29:45 AM »
you really wouldn't need to know that part of it
there would be at least a layer of microcode to isolate the programmer from the physics

it's likely that the first thing we might see is quantum memory arrays
we are talking tera giga bytes - lol
the "hard" drive, as we know it, will become obsolete
you just keep power on a quantum array (if power is required - lol)
and you have fast, reliable, permanent storage

jj2007

  • Member
  • *****
  • Posts: 13958
  • Assembly is fun ;-)
    • MasmBasic
Re: Quantum Progress
« Reply #3 on: October 10, 2012, 07:01:14 AM »
if computers are super-fast, the coders will become super-lazy, and we get super-crappy code

Dave,

Read that article - it's pure Science Fiction. But super-crappy code is a reality, it's called Windows :icon_mrgreen:

mywan

  • Guest
Re: Quantum Progress
« Reply #4 on: October 10, 2012, 08:01:57 AM »
The kind of problems best suited for quantum computing is limited to only a certain class of problems. As such, when such computing does become available, it will be more like just another device attached to a standard computer. Then the processor will simply offload those bits of code best suited for the quantum processor.

News articles tend to massively overplay things.

Gunther

  • Member
  • *****
  • Posts: 4198
  • Forgive your enemies, but never forget their names
Re: Quantum Progress
« Reply #5 on: October 10, 2012, 09:10:53 AM »
Hi mywan,

The kind of problems best suited for quantum computing is limited to only a certain class of problems. As such, when such computing does become available, it will be more like just another device attached to a standard computer. Then the processor will simply offload those bits of code best suited for the quantum processor.

News articles tend to massively overplay things.

that could be, but things are not so easy. What we're doing now with our computers is Aristotelian logic, pressed in silicon - not more, not less. That proposal to realize the  truth tables with switches came from Paul Ehrenfest in 1910. The underlying principle hasn't change since then while the realization of switches changed dramatically (mechanical switches, vacuum tubes, trasnsistors, ICs).

Quantum computers on the other hand use a total other approach. To program such a device (if they ever should succeed) needs a good knowledge about quantum mechanics, especially about the solving of the Schroedinger equation https://en.wikipedia.org/wiki/Schr%C3%B6dinger_equation.

Gunther
You have to know the facts before you can distort them.

Farabi

  • Member
  • ****
  • Posts: 968
  • Neuroscience Fans
Re: Quantum Progress
« Reply #6 on: October 10, 2012, 03:37:26 PM »
this is pretty cool....

http://www.euronews.com/newswires/1684922-french-us-scientists-win-nobel-prize-for-physics/

one downside that i see is that, if computers are super-fast, the coders will become super-lazy,
and we get super-crappy code

If people get a super code, they will stop buying a new computers. I think it is make sense why HLL is so popular these day. Who care about optimiztion except the military.
http://farabidatacenter.url.ph/MySoftware/
My 3D Game Engine Demo.

Contact me at Whatsapp: 6283818314165

anunitu

  • Member
  • *****
  • Posts: 1150
Re: Quantum Progress
« Reply #7 on: October 11, 2012, 02:08:36 AM »
The biggest advantage with quantum computing,as I have read about is that data cannot be hacked in transmission because to copy the data changes it. I don't pretend to understand the physics of this,but it was something I read from an actual Physicist writing on the subject. However it has been said before that something was unhackable,and it got hacked.

Adamanteus

  • Member
  • **
  • Posts: 249
    • LLC "AMS"
Re: Quantum Progress
« Reply #8 on: October 11, 2012, 02:17:40 AM »
Last skeptic postscript : if Nobel committee not mistaken  :dazzled:

hutch--

  • Administrator
  • Member
  • ******
  • Posts: 10583
  • Mnemonic Driven API Grinder
    • The MASM32 SDK
Re: Quantum Progress
« Reply #9 on: October 11, 2012, 11:42:21 AM »
Dave,

Quote
one downside that i see is that, if computers are super-fast, the coders will become super-lazy, and we get super-crappy code

That happened years ago.  :biggrin:
hutch at movsd dot com
http://www.masm32.com    :biggrin:  :skrewy:

dedndave

  • Member
  • *****
  • Posts: 8828
  • Still using Abacus 2.0
    • DednDave
Re: Quantum Progress
« Reply #10 on: October 12, 2012, 05:25:53 AM »
 :biggrin:

Gunther

  • Member
  • *****
  • Posts: 4198
  • Forgive your enemies, but never forget their names
Re: Quantum Progress
« Reply #11 on: October 12, 2012, 06:55:03 AM »
Steve,

That happened years ago.  :biggrin:

I'm not sure: was it before the advent of java or later?

Gunther
You have to know the facts before you can distort them.

mywan

  • Guest
Re: Quantum Progress
« Reply #12 on: October 12, 2012, 11:20:57 AM »
Quantum computers on the other hand use a total other approach. To program such a device (if they ever should succeed) needs a good knowledge about quantum mechanics, especially about the solving of the Schroedinger equation https://en.wikipedia.org/wiki/Schr%C3%B6dinger_equation.

The Schrödinger equation is not particularly important to understand for quantum programming. Basically all you can learn from this is that you "cannot" understand how the logic work mechanistically. At least nobody to date can. Hence there is no requirement that you learn about quantum programming down to a level at which it is by definition incomprehensible. If it was comprehensible, in the sense used here, then it could be modeled (including speed) with a standard computer, but it can't.

To think about it in terms of something resembling Aristotelian logic you have to understand vector math and inner products. If you really want to understand how it's connected to quantum mechanics and observables in everyday life, read Feynman's QED, which is highly comprehensible without any math background. Reading Feynman's QED would be sufficient for me to expand my description of what makes quantum computing so powerful at a fundamental level.

Basically, in a logical sense, you can conceive of a quantum qubit (quantum switch) as a vectorial summation of any arbitrary number of qubits. If you relate it to a unit length, like in Feynman's QED, a qubit that is partly on and partly off (say 60% on) it merely represents an 0.6 unit offset in the sum of the vectors. It also corresponds to a 60% probability the switch will be on when you measure it, since you can't actually measure any single unit vector to be anything but 0 or 1. That's the basic notion behind Schrödinger's cat, with the question being, is the cats state 0 or 1. Once measured you can't return it to its original state before the measurement because the vectors representing this measuring device then get summed over the state of the system that was measured. Though a third measuring device can then sum the probability of the state of the [cat/measuring instrument] state, which remains in a superposition until such a measurement is made.

If you restrict the logic only to unit vectors operations the logic itself is quiet simple. It's only when you try to reconcile its everyday observable consequences with our prejudices about how everyday observables work (via Aristotelian logic) that it blows up in your face. This can be ignored completely, as even most quantum physicist do even within pure science, when speaking in terms of I/O. The term for it is "shut up and calculate". It's justified in the same way you don't object to standard computers on the grounds that switches are never actually off in an absolute sense. Rather their offness exceeded some threshold value. The main difference is that in quantum computing we are dealing with the offness sum of all the switches, rather than each switch individually.

Quantum Programming:
Basically at the machine code level you select qubits the same way you select a memory range, where a register consist of 2 qubits, qureg x1[2] where x1 is the number of registers. Many different quantum programming languages are under development, which some can mix quantum and classical instructions and/or adhere to a classical control flow. You don't even have to obtain a quantum computer to play with programming one. You can simply use (PDF link) Quantum Virtual Machine and program it. The only real difference is it's not going to be fast enough to crack those bank codes to take over the world :badgrin:

mywan

  • Guest
Re: Quantum Progress
« Reply #13 on: October 12, 2012, 12:12:21 PM »
The biggest advantage with quantum computing,as I have read about is that data cannot be hacked in transmission because to copy the data changes it. I don't pretend to understand the physics of this,but it was something I read from an actual Physicist writing on the subject. However it has been said before that something was unhackable,and it got hacked.
The inherit security is a product of theoretical idealizations. This is sound enough as far as it goes, but does not account for the variables involved with any given implementation of it. Feihu Xu, et al has hacked an older commercial QKD system, taking advantage of the noise level threshold allowed by the system. A well written news article is here.

If the attacker is in the position to intercept all of Bobs messages to Alice they can essentially take Bobs transmitted key and resend an entirely different key to Alice. Then upon receiving Alice's messages decode and re-encode them to work with the keys they sent to Alice pretending it was Bobs keys, which they destroyed when they intercepted and read, but Alice doesn't know this so she's using the attackers keys thinking they belong to Bob. Basically two distinct perfectly secure channels, where in one the attacker pretends to be Bob, and in the other channel pretends to be Alice. This would even bypass any of the noise threshold issues in the above article. Though it is possible to defend from this kind of attack it's not explicitly accounted for in the theoretical justification for describing the system as perfectly secure.