|
|
Post by simnettpratt on Apr 4, 2019 1:29:05 GMT -5
So a while back I posted a pic of a quantum computer and it looked like this:
and looked more like a chandelier than the black box I'm used to. I just found out what's going on. The actual quantum chip is in the cylinder hanging low at the very bottom and looks somewhat normal, like this:
Those five horizontal gold disks above are cooling stations. The top one cools the temps to near absolute zero, the next one down cools it a bit more, on down the the final, fifth disk, which cools the lines to 0.015° above absolute zero (the coldest thing in the known universe). At this temperature, the lines become superconductors, and the chip in the cylinder below not only uses no power, but produces no heat. The whole thing is in a big metal closet, with sixteen layers of thermal and radiological protection (it's a big Faraday cage plus).
Currently, we're up to 1,000 qubit CPUs, and the pcs are being used by big players like NASA, Lockheed Martin, and the like. There are two regular servers processing the data, to send it back to NASA or whoever over a regular secure connection. We can add more qubits and it won't draw more power (doesn't draw power now), or get hot (can't get hot).
Not sure what Tomb Raider 87's gonna look like when I get my home quantum pc, maybe like the holodeck on the Enterprise  So the weird chandelier shape is a big cooling device, and what you need to achieve 0.015° K. |
|
|
|
Post by Legend Lover on Apr 4, 2019 2:38:41 GMT -5
That's really interesting. I'm surprised they can get down that low.
|
|
|
|
Post by McWiggins on Apr 4, 2019 3:14:16 GMT -5
Yeah but can it run Crysis?
Seriously though, that is truly amazing!
|
|
|
|
Post by Stearmandriver on Apr 4, 2019 3:51:12 GMT -5
I thought these things were still in concept form. That's incredible!
|
|
Deleted
Deleted Member
Posts: 0
Location:
|
Post by Deleted on Apr 4, 2019 6:18:00 GMT -5
|
|
|
|
Post by simnettpratt on Apr 4, 2019 8:15:21 GMT -5
And if we could get Torvalds to make one it might actually work  |
|
|
|
Post by Cramptholomew on Apr 4, 2019 10:25:24 GMT -5
And if we could get Torvalds to make one it might actually work Didn't Torvalds recently recuse himself from Linux development, or maybe just Kernel development? |
|
|
|
Post by simnettpratt on Apr 4, 2019 13:51:31 GMT -5
What? I hadn't heard that. Am thinking seriously of making a partition for Sparky Gameover Edition (Debian based). It can install several consoles and emulators, and will do Steam and Windows games with Wine and PlayOnLinux. Would be fun to mess with; I've only ever used Gentoo and Ububtu command line distros for servers.
|
|
|
|
Post by sperrytops on Apr 4, 2019 13:59:54 GMT -5
So a while back I posted a pic of a quantum computer and it looked like this:
and looked more like a chandelier than the black box I'm used to. I just found out what's going on. The actual quantum chip is in the cylinder hanging low at the very bottom and looks somewhat normal, like this:
Those five horizontal gold disks above are cooling stations. The top one cools the temps to near absolute zero, the next one down cools it a bit more, on down the the final, fifth disk, which cools the lines to 0.015° above absolute zero (the coldest thing in the known universe). At this temperature, the lines become superconductors, and the chip in the cylinder below not only uses no power, but produces no heat. The whole thing is in a big metal closet, with sixteen layers of thermal and radiological protection (it's a big Faraday cage plus).
Currently, we're up to 1,000 qubit CPUs, and the pcs are being used by big players like NASA, Lockheed Martin, and the like. There are two regular servers processing the data, to send it back to NASA or whoever over a regular secure connection. We can add more qubits and it won't draw more power (doesn't draw power now), or get hot (can't get hot).
Not sure what Tomb Raider 87's gonna look like when I get my home quantum pc, maybe like the holodeck on the Enterprise So the weird chandelier shape is a big cooling device, and what you need to achieve 0.015° K. In ten years we'll each have one of these in our dens. |
|
|
|
Post by simnettpratt on Apr 4, 2019 14:04:56 GMT -5
Maybe. We'll have to have quantum cryptography up and working before the gov will allow us to own one.
|
|
|
|
Post by Legend Lover on Apr 4, 2019 15:44:06 GMT -5
So a while back I posted a pic of a quantum computer and it looked like this:
and looked more like a chandelier than the black box I'm used to. I just found out what's going on. The actual quantum chip is in the cylinder hanging low at the very bottom and looks somewhat normal, like this:
Those five horizontal gold disks above are cooling stations. The top one cools the temps to near absolute zero, the next one down cools it a bit more, on down the the final, fifth disk, which cools the lines to 0.015° above absolute zero (the coldest thing in the known universe). At this temperature, the lines become superconductors, and the chip in the cylinder below not only uses no power, but produces no heat. The whole thing is in a big metal closet, with sixteen layers of thermal and radiological protection (it's a big Faraday cage plus).
Currently, we're up to 1,000 qubit CPUs, and the pcs are being used by big players like NASA, Lockheed Martin, and the like. There are two regular servers processing the data, to send it back to NASA or whoever over a regular secure connection. We can add more qubits and it won't draw more power (doesn't draw power now), or get hot (can't get hot).
Not sure what Tomb Raider 87's gonna look like when I get my home quantum pc, maybe like the holodeck on the Enterprise So the weird chandelier shape is a big cooling device, and what you need to achieve 0.015° K. In ten years we'll each have one of these in our dens. or in our hands. |
|
|
|
Post by simnettpratt on Apr 4, 2019 23:42:20 GMT -5
It will be a long time before we can cool a handheld device to almost absolute zero, and shield it from all radiation. Cramming more and more transistors on a silicon wafer is child's play compared to the cooling and shielding we'd need for an iQuantum |
|
|
|
Post by Legend Lover on Apr 5, 2019 3:38:31 GMT -5
It will be a long time before we can cool a handheld device to almost absolute zero, and shield it from all radiation. Cramming more and more transistors on a silicon wafer is child's play compared to the cooling and shielding we'd need for an iQuantum They could come with their own tank of liquid helium. 😊 |
|
|
|
Post by simnettpratt on Apr 5, 2019 8:19:54 GMT -5
You could wear it on your back, like a flamethrower! Anyway, that would only get you down to about 4°K, and like the gold medal Olympian who's only 4 seconds faster then you, it's a long way to 0.015°K. Plus, you'd have to walk around in a big metal box to get the sixteen layers of shielding. And carry the servers that could send you the data. And they currently cost $15 million to buy, and I don't know what to run. Here's a pic of the big metal closet in the currently working 1,000 qubit quantum computers I referred to in the original post:
|
|
|
|
Post by unknownpipesmoker on Apr 5, 2019 8:29:26 GMT -5
Quantum leap
|
|
Deleted
Deleted Member
Posts: 0
Location:
|
Post by Deleted on Apr 5, 2019 20:09:34 GMT -5
So I only tangentially follow quantum computing but I have a hard time understanding the argument that D-Wave only focuses on one aspect of quantum computing called quantum annealing vs the competitors. Do you have any way of explaining what this is why focusing on one aspect vs whatever everyone else is doing is different in performance? Or maybe it is cost/performance reason? What is the difference in approaches?
|
|
|
|
Post by simnettpratt on Apr 5, 2019 20:57:55 GMT -5
Quantum annealing comes up with a 'good enough' solution, not an absolute one. It could model your plane to fly faster, but not tell you the absolute fastest way there is. This means a quantum annealer can't crack 128-bit encryption, because that requires a definite answer.
A universal gate quantum computer comes up with 'the' answer, so it can crack encryption. It can also be used to solve more equations, and is more useful for more stuff.
Basically, it's the difference between a 'good enough' answer, and the answer. If we can get manufacturing to design reliable and coherent gate qubits, we'll have the computer most folks think of, but we're not there yet; it's still a manufacturing issue. It's not that D-Wave wouldn't like to have a universal gate computer, it's that no one can manufacture a reliable chip yet.
That said, hats off to D-Wave for making the first usable quantum pcs |
|
|
|
Post by simnettpratt on Apr 5, 2019 21:27:07 GMT -5
To better answer your question, what is the difference in approaches, D-Wave had a choice between being a research and development company trying to manufacture a reliable gate pc, and begging for their money, or just make the annealing pcs we can currently manufacture, and selling them for $15 million each, or renting them for goodness knows what an hour.
|
|
Deleted
Deleted Member
Posts: 0
Location:
|
Post by Deleted on Apr 6, 2019 8:15:30 GMT -5
Quantum annealing comes up with a 'good enough' solution, not an absolute one. It could model your plane to fly faster, but not tell you the absolute fastest way there is. This means a quantum annealer can't crack 128-bit encryption, because that requires a definite answer.
A universal gate quantum computer comes up with 'the' answer, so it can crack encryption. It can also be used to solve more equations, and is more useful for more stuff.
Basically, it's the difference between a 'good enough' answer, and the answer. If we can get manufacturing to design reliable and coherent gate qubits, we'll have the computer most folks think of, but we're not there yet; it's still a manufacturing issue. It's not that D-Wave wouldn't like to have a universal gate computer, it's that no one can manufacture a reliable chip yet.
That said, hats off to D-Wave for making the first usable quantum pcs Thank you! This is the answer I was trying to get but nobody seem to lay it out as plainly as you did! |
|
Mac
Full Member
Posts: 834
First Name: John
Favorite Pipe: Ken Barnes Canted Billiard
Favorite Tobacco: Margate, Smyrna, Vintage Syrian
Location:
|
Post by Mac on Apr 6, 2019 9:14:31 GMT -5
The top photo is a model, no? Or a working thing that's not working at the moment? Or summat else?
|
|
Deleted
Deleted Member
Posts: 0
Location:
|
Post by Deleted on Apr 6, 2019 10:35:40 GMT -5
Mac it goes inside the big black box submersed in liquid the computer is used to run parts of the Google search engine. It uses a Linux type program.
|
|
|
|
Post by simnettpratt on Apr 6, 2019 16:53:54 GMT -5
Mac it's just a pic I found to show what they look like. I don't know if the one in the pic is used anywhere, but inside the big metal D-Wave closets in the above pic, you'd see the same gold chandelier/cooling system with the cpu and mobo in the cylinder hanging at the bottom.
|
|
Mac
Full Member
Posts: 834
First Name: John
Favorite Pipe: Ken Barnes Canted Billiard
Favorite Tobacco: Margate, Smyrna, Vintage Syrian
Location:
|
Post by Mac on Apr 6, 2019 19:51:00 GMT -5
OK, thanks, gents. I assumed it had to be tightly enclosed, and it seemed possible a non-working model was made up.
|
|
|
|
Post by simnettpratt on Apr 6, 2019 20:18:45 GMT -5
I don't think they'd drop the money on all that gold alone without intending to put it in a box and trying to get it to run. Gold is currently about $1,300 an ounce, and I believe there's more than a couple of ounces there. I imagine the engineers ain't cheap, either |
|
So the weird chandelier shape is a big cooling device, and what you need to achieve 0.015° K.
