IBM Offering Access To Its Quantum Computer

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
Interested in quantum computing? IBM is offering access to a five-qubit quantum processor to anyone that is interested. Simply sign up here, and then use the web service called "Composer" to create an algorithm by dragging and dropping different quantum logic gates. What are you waiting for? Sign up and give it a try.
 
I'm holding off for seven-qubit. Five just doesn't excite me. :p
 
Alright, I like to think I'm a pretty smart person. I understand the basics of Quantum Mechanics: superposition, waveform probability, quantum uncertainty... But I still don't understand Quantum Computing. Would someone please give me a link that would help me get the intermediate knowledge on this? Not the '7-o-clock news special' 30-second version, I'm willing to take a few hours to read up.
 
Alright, I like to think I'm a pretty smart person. I understand the basics of Quantum Mechanics: superposition, waveform probability, quantum uncertainty... But I still don't understand Quantum Computing. Would someone please give me a link that would help me get the intermediate knowledge on this? Not the '7-o-clock news special' 30-second version, I'm willing to take a few hours to read up.

Heh, I'm right with you. I like (and read up on all the theory) but the computing part I'm remedial at best. A fan of the multiple wave-function collapse theories however far-fetched. How the computing works though, I'd also love the digest version.
 
So do we have any idea how "fast" this thing is? Like, what does it compare to x86 wise? Based on what I've read it actually seems to be a true quantum computer so this has to be like the faster processor in the world, right?
 
You can't ask how fast it is because that's missing the point.

In a conventional processor your clock speed determines the time to answer a question, in a quantum computer the number of operations you can perform determines instead the accuracy with which you can answer that question. Essentially you have some limitation on how long your qubit holds memory before it decays away. You need to perform all your operations on it before that point and retrieve your answer. The number of qubits and the number of operations you can perform are both key parameters. I've talked to my friends over in David Awschlalom's group and the limitation really is apparently no longer the number of operations but simply adding more qubits. It's not like you can just run a wire because the signal you need to carry is a quantum signal. In an electron spin based system that quantum signal is going to be light. They use microwave cavities (Cavity quantum electrodynamics - Wikipedia, the free encyclopedia) to force the light to interact with the electron spin qubit. Then the light can be made to interact with another qubit. If you have any specific questions feel free to ask me, but it's a bit outside my field. I work in nanofabrication for conventional computing devices.

A good ref for quantum algos is here : Quantum algorithms: an overview : npj Quantum Information
 
With research into room temperature superconducting materials like stanene and successful work like this being done on logical processors, it seems like quantum desktop computers are really about 20 years away now. For real this time.
 
You can't ask how fast it is because that's missing the point.

In a conventional processor your clock speed determines the time to answer a question, in a quantum computer the number of operations you can perform determines instead the accuracy with which you can answer that question. Essentially you have some limitation on how long your qubit holds memory before it decays away. You need to perform all your operations on it before that point and retrieve your answer. The number of qubits and the number of operations you can perform are both key parameters. I've talked to my friends over in David Awschlalom's group and the limitation really is apparently no longer the number of operations but simply adding more qubits. It's not like you can just run a wire because the signal you need to carry is a quantum signal. In an electron spin based system that quantum signal is going to be light. They use microwave cavities (Cavity quantum electrodynamics - Wikipedia, the free encyclopedia) to force the light to interact with the electron spin qubit. Then the light can be made to interact with another qubit. If you have any specific questions feel free to ask me, but it's a bit outside my field. I work in nanofabrication for conventional computing devices.

A good ref for quantum algos is here : Quantum algorithms: an overview : npj Quantum Information

reading great posts such as yours has me wondering whether we're going to end up using both old and new skull for a while
 
Yeah, give it a decade +, they might start becoming really viable.

Interesting tech though, some of the current ones are really odd, have been following them a while on the side.
 
Yeah we'll have both old and new for a long while. Quantum computing is great at a lot of specialty things but it's never going to be cheaper to run word on a quantum computer than a classical computer. If anything it will take the place of some supercomputers along with other specialized hardware supercomputers like Anton.
 
Yeah we'll have both old and new for a long while. Quantum computing is great at a lot of specialty things but it's never going to be cheaper to run word on a quantum computer than a classical computer. If anything it will take the place of some supercomputers along with other specialized hardware supercomputers like Anton.

That's true for D-Wave systems which use quantum annealing, but this processor from IBM is a universal quantum processor using quantum gates which has the ability to eventually handle general processing tasks like current processors.
 
Sure it can handle everyday tasks but why would you want it to? Quantum computers offer probabilistic answers and quantum information is not permanent you have perhaps milliseconds before your qubit state decays.

We can make billions of classical gates for 100 dollars which are just as good at a wide variety of operations. We can store billions of bits of classical information. Quantum computers are really ground breaking for some specific problems but your every day classical processor works just fine for typing in office.

I just can't see quantum computers replacing classical within my life time. For mainframe use for simulations or military encryption sure, but not in your home for crysis.
 
Sure it can handle everyday tasks but why would you want it to? Quantum computers offer probabilistic answers and quantum information is not permanent you have perhaps milliseconds before your qubit state decays.

We can make billions of classical gates for 100 dollars which are just as good at a wide variety of operations. We can store billions of bits of classical information. Quantum computers are really ground breaking for some specific problems but your every day classical processor works just fine for typing in office.

I just can't see quantum computers replacing classical within my life time. For mainframe use for simulations or military encryption sure, but not in your home for crysis.

Silicon's pretty much all but done from the looks of it. The only substrates I've seen presented as replacements are 2-D structures which would use quantum mechanics in order to process information.
 
I mean you can use quantum tunneling transistors The Tunneling Transistor - IEEE Spectrum which use a quantum phenomena but still operate in a classical way with classical information. A quantum computer must store and operate on quantum information and that's just not as relevant for a lot of day to day programs. I think classical might hit a soft limit before it goes stupidly parallel but it won't be replaced by a quantum computer anytime soon except in niche specialized research applications where it already exists.
 
Back
Top