An international group of researchers succeeded for the first time in simulating and calculating a chemical process using a quantum computer based on trapped ions
Credit: Harald Ritsch/IQOQI Innsbruck
Simulation of a molecule using a quantum computer
A group of researchers centered on Dr. Cornelius Hempel from the University of Sydney investigated effective ways to model and calculate chemical bonds and reactions using quantum computers. The article was published in the journal Physical Review X of the American Society for Physical Research. "Even supercomputers have difficulty accurately modeling basic chemical processes. A quantum computer can simulate (simulate) nature and reveal new knowledge about the material world. Quantum computers will give researchers new means to solve problems in material science, medicine and the chemical industry with the help of simulations that will simulate what is happening."
The research on quantum computers is still in its infancy and it is not yet clear what problems this device will be able to solve effectively, but the field on which most researchers agree is quantum chemistry. This will probably be the most significant success for quantum computers in the near future.
Quantum chemistry teaches us about the complex chemical bonds between atoms and molecules - these bonds can be studied with the help of the laws of quantum mechanics. With algorithms that can be implemented on quantum computers, scientists hope to discover simpler, more energy-efficient ways to make molecules, so that the materials industry and the medical world can save costs and increase production. Quantum computers may help in deciphering DNA, personalized medicine, increasing the efficiency of solar cells and creating very powerful batteries.
Dr. Hempel, along with colleagues from the Center for Quantum Optics and the Center for Quantum Information in Australia used a quantum computer with 20 qubits to run an algorithm that simulates a chemical bond between a molecule made of the hydrogen atom and lithium hydride (lithium bound to a hydrogen atom). This molecule was chosen on purpose because it is a molecule that is relatively easy to simulate even with classical (non-quantum) computers and thus the scientists can check and compare the result they got with the existing literature knowledge. Dr. Hempel said: "This is a significant step in advancing quantum technology, with the help of this simulation we can examine what works, what doesn't and how it can be improved." In Dr. Hempel's research, he and his team focused on ways to improve the known algorithm for imaging this molecule. By rewriting the quantum code, the researchers found new ways to suppress the errors in the calculation and increase the accuracy in the simulation.
On quantum computers and ion confinement
The information in byte computers is stored and processed in binary form in memory units called bits (smallest memory units) which can be either zero or one. That is, every character, or every command you type on the computer is translated into a set of unique numbers of 0,1. These bits are made of transistors which are actually switches that in certain situations either pass current (in this case we call them 1) or prevent the current from passing through them (in this case we call them 0). In contrast, the bits of the quantum computer (called qubits) are actually ions (charged particles). In order for particles to be used as transistors, a feature must be found that can accommodate at least two different states in the particle (similar to a switch).
This property is called spin - researchers simplify the description of the spin as if the particle rotates around itself just like the Earth rotates around itself and creates the seas. When the particle rotates clockwise its binary value will be zero and counterclockwise its binary value will be one. Unlike the Earth, which can only spin in one direction at a time, a quantum particle can be a "bit" of everything at the same time. This is a well-known property in quantum called superposition that allows the particle to be to some extent rotated clockwise and to another extent to be rotated counterclockwise. This is a non-intuitive feature because it goes against the reality we are used to, but certainly a routine that can be measured in the quantum world. Superposition allows a quantum computer to be in several states at the same time and allow the computer to find the solution to the problem quickly and efficiently, in fact each qubit doubles the number of states the quantum computer can be in at any moment. In order to maintain this feature that is so important to quantum computers (but not the only one) and to perform calculations using them, it is necessary to lock the particles in place and isolate them from the environment.
Using the well-known technology of trapping ions by microwaves, the processor for the quantum computer can be built. The calculation is done by interacting with the particles through focused "shooting" of laser beams at the particles. Even with a very small number of qubits, researchers believe that it is possible to solve problems that no classical computer in the world today or in the future is capable of answering.
2 תגובות
This is an excellent question, the answer to which was not obvious to researchers when they were thinking about quantum computers. The claim that quantum computers can calculate faster should at the end of the day be measured by the number of operations they perform to reach a solution. The smaller the number of operations, the faster the computer. Computer scientists have previously thought about a computer with 3 bit states - 0,1,2 or more and saw that no algorithm in this strange computer could overcome computers with two single states of 0,1, so how a quantum computer, which seemingly contains infinite states for each spin preferable? It turns out that in quantum computers there are algorithms that increase in speed over a classical computer, for example Grover's search algorithm. Let's say you want to search for a single card in a deck of cards. The simple and well-known method is to go through each card separately and check if this is the intended card. If we give the same problem to quantum computers, the number of operations will be at most the square root of the number of cards in your hand.
To actually understand how such an algorithm works, you need to understand a few things about quantum mechanics, but the idea goes as follows - let's say you have 10 cards in your hand and you want to find a single card. You don't know what the card is so you say that anyone can be the card you want with equal probability. We call this claim a quantum state and it really has a physical meaning - let's say the probability is indeed equal, when I tell the quantum computer to draw a card for me, it will randomly choose a card and in such a situation only one tenth of the times it will draw the card we were looking for. From now on we will attach to each card a number that will represent the probability that the quantum computer will pull out the card as soon as it is requested, or in other words when the computer "measures" the system. This is also where the role of superposition comes in - the claim that there is some probability of drawing every card means that my deck is now in a "superposition" state, from which no mixed situation every card in the deck can be drawn.
Now I run on this quantum state (or in our words on the pack of cards) the series of operations of Grover's algorithm - every time such an operation operates on the quantum state the number that is conjugate to the card I am looking for will increase, that is, the probability that the quantum computer will draw it will increase and the other numbers will decrease. In order for the measurement to be sufficiently certain, this operation must be run enough times. The number of times, however, is much smaller than the number of operations on a classical computer, and even becomes very significant when the number of cards is very large (remember that the number of operations to find the card in a quantum computer is the square root of the number of cards, but in a classical computer, on the other hand, it is the exact number of cards there are). So it's true, there may be a situation where the computer will show us a wrong card, but the probability of this will be very small.
There are those who like to claim the following statement about quantum computers: because all the states are related to each other and because they can be in any state at the same time, quantum computers can go through all the possibilities at the same time and reveal the answer to us instantly, compared to classical computers that go through one possibility every moment and therefore it will take They have a long time to come to an answer. This claim is inaccurate - it is true that a quantum computer goes through all the possibilities, but we need to help it to some extent to find the answer we are looking for, and that is why the algorithms exist.
There are problems that a quantum computer does not surpass in its capabilities over a classical computer, and the quantum world is not always an advantage, but in search problems, and other problems a quantum computer has a tremendous advantage.
In the context of the article - quantum phenomena such as chemical bonds in particular encounter these difficulties - because molecules are subject to quantum laws and can be in several states at the same time, the question is how does a molecule decide to manage in this particular way? - In fact, all the particles in nature would prefer to manage in a state that would require the least amount of energy for them to be in, the question is what? There are lots of options to look for and quantum computers can help us find them.
After reading many sources, no one really understands what a quantum computer is.
Everyone knows how to recite what a qubit is, talk about superposition, but no one knows how to explain how it is useful.
If a measurement collapses the state of an ion into a deterministic 0 or 1... what does it help me that in front of you he was in a mixed state?
No one knows how to explain it, and I would be happy if the first wise man appeared here who knows how to explain it in human language.