Slashdot Mirror
First Quantum Computing Gate on a Chip
An anonymous reader writes "After recent success in using quantum computing for superconducting qubits, researchers from Delft have formed the first Controlled-NOT quantum gate. 'A team has demonstrated a key ingredient of such a computer by using one superconducting loop to control the information stored on a second. Combined with other recent advances, the result may pave the way for devices of double the size in the next year or two--closer to what other quantum computing candidates have achieved, says physicist Hans Mooij of the Delft University of Technology in the Netherlands. Unlike today's computers, which process information in the form of 0s and 1s, a quantum computer would achieve new levels of power by turning bits into fuzzy quantum things called qubits (pronounced cue-bits) that are 0 and 1 simultaneously. In theory, quantum computers would allow hackers to crack today's toughest coded messages and researchers to better simulate molecules for designing new drugs and materials.'"
I know grammar has been taking a hit in society as of late, but now even our computers are blatantly spewing out double negatives?
We're not in for an unrough ride, gentlemen.
I find it interesting that the first electronic computing gates devised were the AND/OR gates, using basic diode logic. Quantum computing research develops the NOT gate first. I think this has something to do with the esoteric nature of quantum computing. AND/OR gates require two inputs to change to a single value, where NOT is merely an inverter. The idea of entanglement makes the inversion process a likely first step in quantum research.
For those wondering why this is important, the first true electronic gates were invented in the early 1920's. This predates point-contact transistors by about 20 years, invented in 1947. 60 years later, here we are with transistor computing in every aspect of our lives.
At the rate quantum computing is advancing, I think we can expect to see quantum transistors (in the lab, at least) by 2020. A true useful quantum computer may be available less than 50 years from now. Hopefully by then someone will pick up the slack and have the Linux kernel ported to the Q-CPU architecture!
khasim (12/9/06): In a blind taste test, more people preferred Coke over the Pepsi that I had previously pissed in.
Not Jokes:
It's a Quantum Gate.... NOT!
games journalism blog
but how can they test it when the output is always either 0, meh, pfft or 1?
They whose government reduces their essential liberties for temporary security, receive neither liberty nor security.
I'm no quantum theory expert by a very long shot, but it was my understanding that there are 32 quantum states of electrons, not just on/off (1/0) like in the binary computer world. So, if we now have a quantum NOT gate, doesn't that mean there are 32 possible states of the NOT gate? Also, according to the article the CNOT gates they created can be both 0 and 1 simultaneously. In my mind this would cause errors and actually stop the flow of information instead of speeding it up.
Someone with some understanding of this stuff please elaborate, before my head asplodes.
In theory, quantum computers would allow hackers to crack today's toughest coded messages.
That's an overstatement. A quantum computer will not suddendly magically crack the strongest codes. Yes, certain algorithms designed for quantum computers, like Grover's algorithm, will reduce the time needed to find the key of a symmetrical cipher with about half the number of bits in the key. However, given for example a 256-bit key you would still have ~2^128 keys to check and afaik 2^128 still takes quite sometime to crack....
They're opening the quantum gates now? They're insane! Who knows what might pour out of them... I hope they're at least doing it on the moon.
The future of the human race is up to one lone marine now. Thanks a lot, scientists.
Tribbles & Qubits: The new Pacman
Wanna fight ? Bend over, stick your head up your ass, and fight for air.
At home you will use these for ever more sophisticated rendering of artificially intelligent virtual reality porn.
At work it will be more useful in the advanced simulation of a mechanical process for imprinting letter glyphs on sheets of wood fiber.
Help stamp out iliturcy.
If a qubit is both 0 and 1 at the same time, what is the point of inversing it? Would it then be 1 and 0 at the same time?
Hmm, seems like they've successfully performed a NOT on Moore's law.
Stop! Dremel time!
You'd never know if an article was a dupe or not.
Engineering is the art of compromise.
This is awesome no it's not!
IAMAQP (I am not a quantum physicist) but the theory I read explains a system gaining processing power from shared computing of a single processor replicated across multiple realities. Each qubit is a calculated answer by a machine in one reality and the culmination of those answers assumedly gives you the correct response. David Deutsch wrote a book on this called "The Fabric of Reality" that works through the concept of a basic Turing machine - where computers all come from - and how this can be re-worked into a quantum processor.
There's a lot more math to it than that, but the idea is a simpler approximation formula replicated infinitely across realities gives an accurate response much faster than any single reality calculating the absolute answer.
Cooler yet is that if they're actually making functional quantum gates does this mean the processor power is actually being derived from other realities? That would be awesome and totally Outer Limits material.
-Matt
--- Need web hosting?
Disclaimer: I am a quantum information scientist
Qubits represent a probability of being a 0 or 1. Observing a qubit destroys that probability, and you "read" only a zero or a one.
This is at best an incomplete description of what happens. Qubits are quantum states, not probabilities. Quantum states are sometimes called "probability amplitudes", in that taking the square of the magnitude of the coefficient for a particular basis state gives you the probability of getting that state if you measure in that basis. There are a few very important points: (1) we're dealing with complex numbers, and things work in such a way as to give us the possibility of "interference" of probability amplitudes; (2) quantum states are real states, not just probabilities representing our ignorance of which classical state you'll find when you measure.
A brief intro to the math:
Let's take some qubit in some arbitrary state, which we'll call | psi > (I'm using Dirac notation). We can completely describe the state as follows:
| psi > = a | 0 > + b | 1 >,
where a and b are complex numbers, and have the property that |a|^2+ |b|^2 = 1. We see that we have an uncountably infinite number of possible states for just a single qubit. If psi were a classical bit instead of a quantum bit, we could use essentially the same description, except that the requirement on a and b would then be that exactly one of them is 1, while the other is 0 (only two possible states). If psi were a "classical" analog "bit" or a probabilistic bit, the requirement would be that a, b in [0,1], and a+b=1.
What happens if we measure psi? It depends on the basis we choose to measure in, but if we go to measure psi in the {| 0 >, | 1 >} basis, we'll get | 0 > with probability |a|^2, and | 1 > with probability |b|^2. Figuring out probabilities for other bases requires only a basis transformation (simple linear algebra).
Now, this qubit business seems horribly messy--we have an infinite number of states for a single qubit--how can we possibly describe the action of a two-qubit gate like controlled-NOT (CNOT)? Fortunately, quantum mechanics is linear, which means that if we describe how a gate operates on each of the possible input basis states, we've completely specified the gate. For two qubits, we can use the following basis: {| 00 >, | 01 >, | 10 >, | 11 >}. Labeling the rows and columns in that order, we get the following truth table for the CNOT gate:
1 0 0 0
0 1 0 0
0 0 0 1
0 0 1 0
In other words, if the first bit is 0, do nothing to the second bit. If the first bit is 1, flip the second bit.
It turns out that CNOT plus a bunch of different single qubit gates is universal, meaning you can use that set of gates to implement any "quantum circuit".
here's a simple representation of a C-Not on two qbits:
:= in_x; := ((not)in_y and in_x) or (in_y and (not)out_x)
______
-| NOT|-
-|____|-
the gate has two inputs and two outputs
in direct computation the the inputs are on the left, the outputs on the right
in reverse computation the other way around
let's take direct computation,
say the inputs(left side) are in_x(top) and in_y(bottom) and the outputs(right) out_x(top) and out_y(bottom)
the C-Not performs the following function:
out_x
out_y
BUT it performs this functions simultaneously on a superposition of inputs,
a superposition of inputs for a qbit roughly translates to: something like if I measure the qbit I may get 0 with p probability and 1 with (1-p) probability
algebraically you would express this using in_y = sqrt(p) |0> + sqrt(1-p) |1>
so literally the gate works like this: in the cases when in_x is 1, out_y will be 1 with a probability of p and 0 with a probability of (1-p) - exactly the inverse probabilities of in_y
this is all roughly speaking, there are other more subtle aspects...