Comments on: How to build a million-qubit quantum computer http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer Accelerating Intelligence Thu, 20 Jun 2013 09:23:34 +0000 hourly 1 http://wordpress.org/?v=3.4.1 By: libra9 http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-84242 libra9 Sat, 05 Jan 2013 17:52:28 +0000 http://www.kurzweilai.net/?p=172881#comment-84242 Sorry, that is "tim" the realist. Sorry, that is “tim” the realist.

]]> By: libra9 http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-84241 libra9 Sat, 05 Jan 2013 17:51:16 +0000 http://www.kurzweilai.net/?p=172881#comment-84241 I am impressed with the sophistication and civility of the dialogue between GAUSS and time the realist. Perhaps there IS some hope for mankind. I am impressed with the sophistication and civility of the dialogue between GAUSS and time the realist.

Perhaps there IS some hope for mankind.

]]> By: GAUSS http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-63574 GAUSS Thu, 06 Dec 2012 18:02:18 +0000 http://www.kurzweilai.net/?p=172881#comment-63574 Ok, I see. You're essentially describing quantum Monte Carlo methods. It could work, you'd just never know if you had *the* optimal solution unless you had a purely deterministic quantum computer (functionally 100% accurate). Even so, it's a great idea, and one worth exploring. :) Ok, I see. You’re essentially describing quantum Monte Carlo methods. It could work, you’d just never know if you had *the* optimal solution unless you had a purely deterministic quantum computer (functionally 100% accurate). Even so, it’s a great idea, and one worth exploring. :)

]]> By: tim the realist http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-63408 tim the realist Thu, 06 Dec 2012 13:27:00 +0000 http://www.kurzweilai.net/?p=172881#comment-63408 You are missing the point. Traveling slaesman is exactly the type of problem that only requires a 1% accuracy ot be successful. Only run 10,000 times on a quantum computer and pick out the best path it comes up with. With 10,000 answers from the quantum computer with a 1% chance any one of them is actually correct there is only a 2.24877E-44 probability you won't have the absolute best answer contained within the answer set. Going up to 1,000,000 or so runs lower the probability so low that it basically can't happen. This is close enough for the best solution from the data set to be used as the actual solution for these type of problems You are missing the point.
Traveling slaesman is exactly the type of problem that only requires a 1% accuracy ot be successful. Only run 10,000 times on a quantum computer and pick out the best path it comes up with. With 10,000 answers from the quantum computer with a 1% chance any one of them is actually correct there is only a 2.24877E-44 probability you won’t have the absolute best answer contained within the answer set. Going up to 1,000,000 or so runs lower the probability so low that it basically can’t happen. This is close enough for the best solution from the data set to be used as the actual solution for these type of problems

]]> By: GAUSS http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-63213 GAUSS Thu, 06 Dec 2012 07:17:05 +0000 http://www.kurzweilai.net/?p=172881#comment-63213 Ex: Traveling Salesman Problem. A quantum computer would likely be very good at this, but it would take most classical computers on the order of several million years to check a calculation. There are proofs for this, but I won't invoke them here. ;) Ex: Traveling Salesman Problem. A quantum computer would likely be very good at this, but it would take most classical computers on the order of several million years to check a calculation. There are proofs for this, but I won’t invoke them here. ;)

]]> By: YeahYeahYeah http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-62951 YeahYeahYeah Wed, 05 Dec 2012 22:57:09 +0000 http://www.kurzweilai.net/?p=172881#comment-62951 Anthrobotic, this guy disagrees with you: http://www.technologyreview.com/view/508246/why-the-universe-is-not-a-computer-after-all/ Anthrobotic, this guy disagrees with you: http://www.technologyreview.com/view/508246/why-the-universe-is-not-a-computer-after-all/

]]> By: tim the realist http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-62851 tim the realist Wed, 05 Dec 2012 19:55:03 +0000 http://www.kurzweilai.net/?p=172881#comment-62851 Please provide an example where a classical computer cannot verify if one proposed solution to a computational problem is "better" than another. "Better" may mean more precise, shorter path, lower cost, etc. Usually the number of possible solutions to an NP problem is very large and requires checking all of them to find the best. Using a low reilability quantum computer to provide a set of possible answers to check can quickly reduce the size of the solution space. Please provide an example where a classical computer cannot verify if one proposed solution to a computational problem is “better” than another. “Better” may mean more precise, shorter path, lower cost, etc.

Usually the number of possible solutions to an NP problem is very large and requires checking all of them to find the best.

Using a low reilability quantum computer to provide a set of possible answers to check can quickly reduce the size of the solution space.

]]> By: GAUSS http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-62719 GAUSS Wed, 05 Dec 2012 16:14:52 +0000 http://www.kurzweilai.net/?p=172881#comment-62719 After a certain point, you wouldn't be able to use classical computers to verify any answers. The problems, especially in discrete optimization, just get too big for any i7, GPU, or otherwise. :( After a certain point, you wouldn’t be able to use classical computers to verify any answers. The problems, especially in discrete optimization, just get too big for any i7, GPU, or otherwise. :(

]]> By: tim the realist http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-62594 tim the realist Wed, 05 Dec 2012 13:13:44 +0000 http://www.kurzweilai.net/?p=172881#comment-62594 Can anyone explain to me why reliability is so important in these types of computing architectures? My understanding is that this type of computer would be used primarily for NP type problems. Why can't these architectures simply be scaled up at the cost of reliability and have the program run multiple times. If it's right ~1% of the time, then that should be plenty good enough to use classical computers to analyse thousands or millions of result states to see if they are any better than standard classical solutions. Factoring large numbers for instance is easy to check potential answers. So if you get hundreds or thousands of potential answers they can all be quickly checked using classical architecture to identify the final correct solution. Many NP problems are like this. Can anyone explain to me why reliability is so important in these types of computing architectures? My understanding is that this type of computer would be used primarily for NP type problems.

Why can’t these architectures simply be scaled up at the cost of reliability and have the program run multiple times. If it’s right ~1% of the time, then that should be plenty good enough to use classical computers to analyse thousands or millions of result states to see if they are any better than standard classical solutions.

Factoring large numbers for instance is easy to check potential answers. So if you get hundreds or thousands of potential answers they can all be quickly checked using classical architecture to identify the final correct solution. Many NP problems are like this.

]]> By: anthrobotic http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-62075 anthrobotic Wed, 05 Dec 2012 00:52:55 +0000 http://www.kurzweilai.net/?p=172881#comment-62075 Apologies, mistaken link in comment #2. For the correct link to a simulation theory piece with a dusting of quantum mechanics/computing flavor: http://goo.gl/ZD9cJ Apologies, mistaken link in comment #2. For the correct link to a simulation theory piece with a dusting of quantum mechanics/computing flavor: http://goo.gl/ZD9cJ

]]> By: anthrobotic http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-62072 anthrobotic Wed, 05 Dec 2012 00:50:38 +0000 http://www.kurzweilai.net/?p=172881#comment-62072 AHHHHHH!! WRONG LINK!!! Apologies. For the correct link to a simulation theory piece: http://goo.gl/ZD9cJ AHHHHHH!! WRONG LINK!!!
Apologies. For the correct link to a simulation theory piece: http://goo.gl/ZD9cJ

]]> By: eldras http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-61881 eldras Tue, 04 Dec 2012 18:44:26 +0000 http://www.kurzweilai.net/?p=172881#comment-61881 Planck (inventor of quantum theory) was a crackpot who believed there was a divine intelligence controlling everything. we dont know much about QT and are using trial and error a lot. Planck (inventor of quantum theory) was a crackpot who believed there was a divine intelligence controlling everything. we dont know much about QT and are using trial and error a lot.

]]> By: GAUSS http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-61824 GAUSS Tue, 04 Dec 2012 17:23:00 +0000 http://www.kurzweilai.net/?p=172881#comment-61824 Not a cat lady! Noooooo! Not a cat lady! Noooooo!

]]> By: anthrobotic http://www.kurzweilai.net/how-to-build-a-million-qubit-quantum-computer/comment-page-1#comment-61756 anthrobotic Tue, 04 Dec 2012 15:43:07 +0000 http://www.kurzweilai.net/?p=172881#comment-61756 With advances in quantum computing, possible "pixelation" of the universe, and some supremely annoying probability, seems reality most likely is a computer simulation. God, an intensely needy & cosmically powerful cat lady, is just playing the latest release of "Semi-Conscious Hairless Primates." Yep. More: http://goo.gl/2wkcN -Reno at Anthrobotic.com With advances in quantum computing, possible “pixelation” of the universe, and some supremely annoying probability, seems reality most likely is a computer simulation.

God, an intensely needy & cosmically powerful cat lady, is just playing the latest release of “Semi-Conscious Hairless Primates.”

Yep.
More: http://goo.gl/2wkcN

-Reno at Anthrobotic.com

]]>