Occasionally, as everyone in the scientific and technical community clings to a dubious degree of calm, one happens to come up with a new innovation that turns things upside down, changes the rules that players follow in those circles, and opens the door to a horizon that nobody would have seen before. The most famous American philosopher of science, Thomas Kohn, calls it the “scientific revolution,” and we seem to be facing one of those interesting and important uprisings at the same time, according to a recent announcement by Google.
Undoubtedly, the most exciting technical news of the past few days has been Google's announcement of "quantum supremacy". One of its newly developed quantum computers was able to address a computer problem that takes on the most advanced computers of 10,000 years. It only takes 200 seconds for Google to complete it, called Sycamore, according to a new paper published several days ago in the prestigious journal Nature.
Quantitative superiority
Quantum supremacy simply means the ability of a quantum computer to perform tasks that are impossible to do by the most capable of modern computers. In fact, the problem that Sycamore manages can be dealt with by classical computers as well. The period - 10,000 years - for a smaller period, but the difference will always be large between them and Sycamore, so our little friend has become, according to the statement of Google company, the first supercomputer in history!
Surprisingly, this vast difference between the two periods undoubtedly leads you to wonder about the mystery of the game. How did Google's computer manage to make that huge leap? At that point, let's go back to your calculator, the one you used at school to do some calculations, print out 158, then hit the "x" and print 169 to show the result immediately. Let me try it now, yes, it's 26720, very fast, right? You will need more than a minute or two to solve it with a pen and paper if you are good at calculation, but how does the calculator perform that process so easily ?!
In your normal life you can use ten numbers to express any other number, it's numbers from zero to 9, if you put one next to another one you get 11, and if you decide to write a million and six hundred thousand you will put one and to the right of six and then five zeros. For a computer, that luxury doesn't exist. All our electronic friend has two numbers, one and zero, but you can still use them to write all the other numbers, by simple encryption, it's what we know as the "binary counting system". For example, we can refer to zero as 0000, one to 0001, two to 0010, etc., and so on for all numbers.
If we encode the basic numbers into that system of 1 or 0, then we can encode everything else, from images to audio, images are made up of very small pixels, each with a color tone that can be given a number, and for each part A small amount of sound can give you a number too. In this universe, we can replace anything with a number. What you need now is to identify the computer with those codes and will know the numbers, but how does the computer know that?
Torrent of signals
The idea is simple, and depends on electrical signals. If you have a copper wire, we can consider the passage of an electrical signal to it as information that means "1" and that the current does not pass "zero". This is the idea of transistors in computer processors. Do not pass it, thus giving the device a 1 or 0. The whole computer unit, called "bit," works so that we could give the calculator the number data we just talked about if we had four bits, each of which indicates 1 or 0.
These electrical signals then enter what we call "logic gates" (5), which are simply a set of transformers that perform certain functions. Some gates, for example, reflect the signal, the current (1) enters the gate to exit (zero), or vice versa, others It may enter two signals and, for example, reverse one of them if the other is a particular number, or produce a signal (0) if any of the next two signals (0), these gates are simply the same basic logical processes that form the foundations of mathematics, which can stretch So we can do the collection.
If you can do a collection, you can do the multiplication, simply because "5 * 6" means that we can add five times six, and if you can do the basic calculations, you will be able to do anything else, and theoretically the computer can do any calculation but The problem is in time and capacity, to understand the idea, let's assume we have a computational problem that we would like the computer to solve, and says that there are four people we want to distribute to two cars, but two conditions, the first is to put people who love each other together in the same car, and the second to put Who hate each other in separate cars. The idea is that Ali loves Salma, B "Hussama", but there is no relationship between this duo and the other duo.
At first glance you will think that this is a simple process, put Ali and Salma in a car, put Nabila with Hossam in another car, which is already so for the computer, it is first to try all the possibilities, and will be 16 probabilities here, and then by playing signals through logic gates up to the solution It is an easy process that a simple computer can answer in much less than a second, simply because the flow of the current represented by the signals - in the processor - is close to the speed of light, but what if there were 100 people, not three? What if there were a trillion people? Every time we double the odds, we only have two cars, so the computer will have to examine each one individually.
Quantum overlay
Well, now we are in direct confrontation with Sycamore, but we must first get to know the quantum world, since one of the most famous properties of subatomic particles is that they can exist in two cases at the same time, it is what we call quantum superposition (6) To understand it, imagine that, with a finger stroke, we wrap a coin representing one Egyptian pound on a wooden table.While a coin is rotating, we don't know in what way a king or a writing will dock here. All possible possibilities together. "
The state of quantum superposition is similar, where an electron can, for example, be found in two different states of winding (top and bottom) at the same moment. Spin is a property of subatomic particles, which is like Earth orbiting itself. It is not exactly the same; it is only the earliest example we know with an important difference: the Earth orbits because it was pushed to rotate at the beginning of its life in the solar system.
We can easily recognize these two states of winding, and transistors can be made to differentiate between the two states, which means that we can use them to represent the computer language (zero and one), but the state of quantum composition allows the particle to exist in both cases. , Which means that each bit (called QBit (8) in the case of Quantum) can hold two probabilities at the same moment, and if we have 4 bits, the odds are multiplied by 16 operations at the same moment, 16 probes consecutively, in this case it can be examined all at the same moment, because Each bit has all the possibilities together.
This is what organs like Sycamore do and hence its power. It does not calculate sequentially, but in parallel, in the sense that you can say that at some point if we decide to assemble the world's most powerful organs, we can reach that capacity, but you do not yet understand the secret. The power of quantum computers, where power rises so exponentially in the case of Qbits, that only 300 Qbits can process more than the number of particles of the entire universe! The secret of quantum computers is therefore not the speed of the signals, they all run at the same speed between the transistors, but in the way they sequence, the ordinary computer goes through them one by one, the quantum computer goes through them, and when we go up the possibilities up, every new "qubit" not only adds For the process;
Shall we emulate all those things?
At that point you might say, "Why don't they do that as long as it's that simple?" The answer is that, unfortunately, the nature of the quantum world is likely to be printed, meaning that the equation that velocity equals distance over time does not easily apply there below. Not a fixed number, but a probability, for example - let's go back to the previous currency example - when you put your hands on the revolving currency, you are likely to find a "king" by 50%, and find "writing" by 50% as well. The quantum computer brings it to you is one possibility, because we can't detect an electron in a quantum T immediately quantum mode to normal mode (Classic).
This is a big problem, because when you do a calculation you want to get the most accurate answers, what if the probability of that process a second, third or fourth answer but not the best? At that point there is a major quantum conflict between companies, where we must first tamper with the mechanism by which the probabilities come out of what we call the wave function, so that we ensure that it will go to the path of the correct answers, That. Well, Sycamore is therefore not just a 53-qubits ordinary computer, but a masterpiece of such magnitude that it can handle an estimated volume of information (2Q53).
Immediately after the launch of the scientific paper of Google (9) "IBM" that we must look at these results with a degree of skepticism, pointing out that its computer, "Sumit", which fills an area the size of two basketball courts, at Oak Ridge National Laboratory in Tennessee, The same calculation in 2.5 days and not 10,000 years, but don't you notice that jump? Assuming the truth of IBM's claim, there is still a huge difference between 200 seconds and two and a half days!
Google announced in a post that it will work in the next phase on two separate plans, the first involves inviting researchers in several domains to start using Sycamore in an effective manner, as what he has done so far is not useful in the sense understood outside the realm of quantum and requires further development in algorithms To give real production with controlled possibilities, Google will also invest in quantum computer technology so that in the future it can better adapt to possible failures.
At that point, the enthusiasm must be curbed a bit, as there are still many problems of quantum probability control, as well as other problems related to the development of logical algorithms and interfaces that can deal with this new state. All we have done in this report is to simplify the crusts So that our reader can digest some of the general idea.
It was like tonight yesterday, in the early days of ordinary computers where a simple calculation was required to make a device the size of a whole room and wait a few hours for a result that might not be correct. At the time it was useless, but it was a beginning that allowed us to evolve. right Now. If you decide to ponder a researcher's image while tweaking a computer's connections more than half a century ago, and in return you decide to ponder a researcher's image now, you will be amazed at the degree of similarity, because in fact everything starts like this, large, complex and surrounded by a majority that says It will not fit and you are wasting your time, then here you are. This model applies to almost everything from computers to your life.
Have we come close to a computing power that will finally enable us to make a precise simulation of the human brain, of proteins in our bodies, of extremely distant galaxies or the beginnings of the universe, of artificial intelligence capable of simulating human emotions and behavior to complicate it? The higher the ability of computers to process, the higher our ability to probe the unknown, which means that what Google did in its recent paper is not just a primitive computer operation that is not useful in a sense, but it is a revolution that can change everything else in our lives, we just have to calm down. A little, and do not stop working.