# Does processing for a quantum computer take place in other universes?

Apologies in advance if my question seems misinformed. I am a software developer, and neither quantum mechanics nor physics are my specialties.

The exact physical mechanism at work within the quantum computer is somewhat theoretically complex and intuitively disturbing. Generally, it is explained in terms of the multi-world interpretation of quantum physics, wherein the computer performs calculations not only in our universe but also in other universes simultaneously, while the various qubits are in a state of quantum decoherence.

The amazing power of the quantum computer stems from the fact that if you have a collection of qubits – a register – in which each qubit is in an indeterminate state, then the register effectively represents all possible numbers at once. If you then perform a single computation on the register, the computation works on every possible number, simultaneously. David Deutsch explains the process using the idea of parallel universes – although we see only the single register in our universe, it actually exists in many other universes too, one for each of its possible states. By operating on the register in our universe, we kick off computations in all the other unseen universes, and then magically retrieve the answer.

Assuming the multiverse interpretation of quantum mechanics is correct, does this mean that a quantum computer is capable of considering all possible solutions to a problem simultaneously because it delegates the consideration of (calculations for) each possible solution to a different universe?

-

The idea is somewhat far fetched, but a good one. Did you know that a quantum computer that can run at room temperature has been built? (Only for 39 minutes though) Well to answer your question: Quantum computers make use of the theory in quantum mechanics that nothing is defined, and everything can exist in multiple states. For example, quantum mechanics predicts that a cat can be both dead and alive at the same times. Difficult concept to grasp, yet that's what it says. Quantum computers, unlike the ordinary computers, have 'quibits' which can exist as 0, 1, and a superimposition of 0 and 1 (hence it can exist as both 1 and 0 at the same time). There are a total of $2^n$ states that a quibit can exist in.

So what is it that makes quantum computers so fast? They can solve a problem requiring 100 steps, thus 100 bits in the normal computer, in 1 step, using one 'quibit' as that one quibit can exist in all those 100 states at the same time. Another fundamental property is that quantum computers use complex algorithms to solve even the most basic of questions. Those algorithms, rather being operated by the software of the computer, are the software of the computer. Those algorithms are what the computer is based on. The algorithms never reach the ultimate answer to a question, as quantum mechanics says that there is no definite answer to a question. Rather it calculates an answer with the highest probability and takes that to be the correct answer.

Also remember, that quantum mechanics itself doesn't predict the existence of multiple universes, rather the theory of multiple universes has been created in order to interpret the quantum mechanics theory. There is no need to say the calculations exist in multiple universes, as quantum mechanics says that for example, that an electron can exist in two places at once in our universe. Same can happen with a quibit, it can exist in 2 states simultaneously, thus greatly reducing the number of steps and the time required for calculations.

-
"...predicts that a cat can be both dead and alive at the same times." To my knowledge, Schrödinger's cat was suggested as an apparent paradox, illustrating the shortcomings of the Kopenhagen-interpretation of QM, especially when applied to macroscopic objects. The notion of 'observation/measurement' is really not well defined. Isn't the cat an observer itself? A way out could be quantum decoherence, by which the superposition is lost during time evolution due to interaction with a suitably large environment. – Nephente Jun 25 '14 at 7:04
Well yes... But it's an example easy to understand and helps explain the basis of the theory? Does it not? @nephente – Gummy bears Jun 25 '14 at 15:37
In my opinion it rather demonstrates the pitfalls when trying to naivly apply qm to macroscopic objects. Especially the layman is left with the impression, that the randomness inherent to QM is directly applicable to the macroscopic world. Worse, it leads to the believe that somehow the human observer plays a distinct role in the workings of nature. That reality depends on the act of a concious being conducting an observation. I guess the physicist should take it as an example that 'measurement' is ultimatly not well defined, although it is a useful concept. – Nephente Jun 25 '14 at 16:55
Assume the cat is an inanimate object then. You really can't observe an electron that easily, can you? That is not the issue at hand, we are not discussing the workings of quantum mechanics, much of which continues to be theoretical. It helps to demonstrate a point which is necessary to help understand the question, and this example is the most obvious and plainest to see. Of course you cannot apply qm to macroscopic objects, however, it is difficult to understand it without such an example. I believe that it is more important to answer the issue at hand, rather than worry about workings of qm – Gummy bears Jun 25 '14 at 16:58

"Assuming the multiverse interpretation of quantum mechanics is correct, does this mean that a quantum computer is capable of considering all possible solutions to a problem simultaneously because it delegates the consideration of (calculations for) each possible solution to a different universe?"

This is close to the truth but there are important caveats. A quantum computer can put a register in a state in which if you measured the observable in which the outcome is being stored there would be multiple versions of you after the measurement and each version would see a different outcome. If you actually measure the register this way then the computation won't work. Getting the right outcome depends on information that is only present in how all of the different versions are related to one another. If you spread that information around then it is no longer present in the register alone and no operation on the register alone will produce the right answer so the computation won't work. So you only get the right answer if you don't differentiate into multiple versions, one for each value of the register.

Also, since getting the right answer involves multiple versions of the register interacting it is wrong to say that the register differentiates into parallel universes: what happens inside the computer is more complicated than that, see

See also "The Fabric of Reality" and "The Beginning of Infinity" by David Deutsch.

-

It sounds like you are asking whether we can clone the qubits during the computation without affecting the original qubits. This would be useful but has been proved to be impossible. It is called the No Cloning theorem and is fundamental to quantum mechanics. No interpretation of quantum mechanics can circumvent it.

-