InterviewQuantum Leap: "The big bang of quantum computing will come in this decade"
Interview
Quantum Leap: "The big bang of quantum computing will come in this decade"
Prof. Dario Gil, head of IBM Research, is at the forefront of the quantum computing revolution, which is about to fundamentally change our world. In an exclusive interview, he explains what are the challenges that have delayed the realization of this dream, why you need to be an Einstein to understand it, and when we’ll finally start using it ourselves
In the few images that IBM has released, its quantum computing lab looks like the engine room of a spaceship: bright white rooms with countless cables dangling from the ceiling down to a floating floor, pierced with vents. This technological tangle is just the background for the main show: rows of metal supports on which hang what look like... white solar boilers.
There, within these “boilers”, a historical revolution is taking shape. IBM, a computing dinosaur more than a century old, is trying to reinvent itself by winning one of the most grueling, expensive and potentially promising scientific races ever: the race to develop the quantum computer. "We are living in the most exciting era in the history of computing," says Dario Gil, Senior Vice President of IBM and head of the company's research division, in an exclusive interview with Calcalist. "We are witnessing a moment similar to the one recorded in the 40s & 50s of the last century, when the first classic computers were built." A few weeks after this conversation, his statements were further confirmed, when the Nobel Prize Committee announced the awarding of the prize in the field of physics to three researchers whose research served as a milestone in the development of the field.
The name Dario Gil shakes a lot of quanta and cells in the brains, and maybe even in the hearts, of physicists and computer engineers all over the world. This is the person who leads the most advanced effort in the world to develop a quantum computer. In September, when Gil landed in Tel Aviv for a short visit to give the opening lecture at the IBM conference, the hall was packed with senior engineers, researchers from the top universities in Israel, and representatives of government bodies - all enthralled by what Gil had to say.
Gil (46) was born in Spain and moved to the United States to study at MIT University. He completed his doctoral studies there, and immediately after graduation began working at IBM in a series of research and development positions. Since 2019, he has been leading the company's research division, which has 3,000 engineers at 21 sites, including Israel. Under his management, in 2016, IBM built the first quantum computer whose services are available to anyone: if you have a complicated question, you can go to the IBM Quantum Experience website, remotely access one of the quantum computers through the cloud - and, perhaps, receive an answer. But as with everything related to quantum computing, it just sounds simple.
"Quantum computing is not just a name for an extremely fast computer," says Gill. In fact, he explains, the quantum computer is no longer a supercomputer that uses the same binary method that is accepted in every classical computer, but a completely new machine, another step in the evolution leading from strings of shells, through beaded invoices and calculating bars, to gear-based mechanical computers, to the electronic computer— and now to the quantum computer. "Essentially, the quantum computer is a kind of simulator of nature, through which it is possible to simulate natural processes, and thus solve problems that previously had no solution," explains Gil. "If the classical computer is a combination of mathematics and information, then quantum computing is a combination of physics and information."
This connection makes it possible to solve certain types of problems with unprecedented speed: Google, which is also developing a quantum computer, claimed in 2019 that it had reached "quantum supremacy" — a demonstration of a calculation that a quantum computer would perform more efficiently than a classical computer. The researchers at Google showed how a quantum computer performed in 200 seconds a calculation that they claim would have required a classical computer ten thousand years to complete. This claim has since been disproved by other researchers, who have presented an algorithm that allows a classical computer to perform the same calculation in a reasonable amount of time—but even this Google failure provides an idea of the enormous power a quantum computer will have.
"The quantum computer does not make the classical computer superfluous: they will live together, and each of them will solve different problems," explains Gil. "It's like asking you how to get from point A to point B: you can walk, ride a bicycle, travel by car or fly. If the distance between these points is 50 km, you won't fly between them, right? Accordingly, it is a mode suitable for a classic computer. A quantum computer allows you to fly, even to the moon, and quickly."
You will soon explain to me how it works, and in which areas exactly, but before that, let's start from the bottom line: what can we do with it?
"Quantum computing will make it possible to crack a series of problems that seemed unsolvable, in a way that will change the world. Many of these issues are related to energy. Others are related to the development of new and exciting materials. We tend to take the materials available to us for granted, but in the past there were eras that were defined by the materials that dominated them - ‘The Stone Age', the 'Bronze Age', the 'Iron Age'. Quantum computing will help us develop materials with new properties, therefore the first sector that is already using it is industry, especially the car industry: the car manufacturers are interested in better chemistry, which will enable the production of more efficient and durable batteries for electric vehicles. For a normal computer this is a huge task, and to complete it we have to give up accuracy and settle for approximate answers only, but quantum computing can help quickly develop materials that will fit the task, even without entering the lab. The efficiency of a quantum computer when it comes to questions in chemistry is also used in the pharmaceutical industry, There they are beginning to make initial use of such computers to examine the properties of molecules, and in this way to speed up the development of new drugs; and also in the fertilizer industry, which will be able to develop substances whose production will not harm the environment.”
The uses are not limited to the material world. "For the financial sector, for example, the quantum computer enables the analysis of scenarios, risk management and forecasting, and the industry is already very interested in such possible applications, which could provide the general public with dramatically improved performance in investment portfolios,” for example.
At the same time, there are industries that quantum computing will force to recalculate their course, and the information security industry is at the forefront. The modern encryption systems (mainly RSA, one of whose developers is the Israeli Prof. Adi Shamir) are asymmetric: each recipient publishes a code that allows the information sent to them to be encrypted ("public key"), which includes the product of two large prime numbers that are kept secret. To decipher the encrypted information, this product must be broken down into factors - but without knowing what the initial numbers are, "this task would require a normal computer to calculate for many years," explains Gil. "However, for the quantum computer, such a calculation can be a matter of seconds."
There is a real threat here to an entire industry, the logic behind which has been built since the 1970s, and now suddenly the ground is cracking under it.
"True, a normal computer needs ten thousand years to solve an encryption that a quantum computer would solve in an instant. That is why the quantum computer threatens the world of cyberspace and encryption, which are the basis of all global information security. This is an example that is not related to physics or nature, but simply to the stronger and faster computing power of the quantum computer.”
The computer that works against all the rules of intuition
To understand the power of the quantum computer, this concept, "quantum computing", must first be broken down. The first step is to stop thinking in the familiar concepts of one and zero. Forget about bits and binaries. The key to understanding quantum computing is the recognition that this dichotomy is not there: instead of the bit, quantum computing relies on a basic unit of information called a qubit (short for "quantum bit"). The qubit is simultaneously one, zero and everything in between.
This is the moment to stop and explain the theory that underlies the quantum computer, and which seems to go against common sense. "Quantum theory makes it possible to explain the behavior of very, very small particles," Gil explains. "At school we are presented with a model of an atom that looks like a planet, with a nucleus and electrons moving around, but at the beginning of the 20th century, this model turned out to be not very accurate." This happened when physicists such as Max Planck and Albert Einstein realized that light, which until then physics saw as a wave, also behaves as a particle - and the energy of this particle can only be described in "quantum" jumps, that is, as discrete packets. In the decades that followed, this theory was developed more and more, and proved to be effective in describing a variety of phenomena in the world of particles. And yet, its deep meanings remain obscure even today.
Such is, for example, the idea that a particle is in more than one place. According to quantum theory, a particle moving between two points moves simultaneously in all the paths between them, a state called "superposition". It's not that we don't know its exact location: it just doesn't have one. Instead, it has a distribution of possible locations that coexist. In other words, reality is not certain, but probabilistic.
And this is not the only puzzle posed by quantum theory. Another confusing concept is "entanglement", a situation in which several particles exhibit identical physical values, and respond simultaneously to a change in one of them, even if they are at a great distance from each other. Gil suggests thinking of it as tossing two coins: anyone who has studied statistics knows that the probabilities of getting a "head" or a "tail" on each of them are independent. But in the quantum model, if the coins (representing particles here) are intertwined, then tossing one of them will result in the same result in the other. "Einstein didn't believe in interweaving, and hated these patterns," Gil says with a smile.
Measurements that affect the results? A reality that is not absolute but statistical? Particles that become twins even at infinite distance? If these ideas sound puzzling, incomprehensible or counter-intuitive to you, you are not alone: "Whoever comes across quantum theory and is not left stunned, has not understood it," said the physicist Niels Bohr, Einstein's contemporary and his great nemesis, who won the Nobel Prize for his contribution to the development of the theory (Einstein, by the way, had reservations about Bohr's interpretation of the theory's conclusions). Another physicist who won the Nobel Prize for his contribution to the theory, Richard Feynman, commented on this when he said: "If you think you have understood quantum theory, you have not."
The same Feynman is the father of quantum computing: he wanted to simulate the behavior of particles, but due to the probabilistic nature of the theory, a classical computer that would try to perform such a simulation would require an enormous amount of calculations, so that the simulation would become impractical. "Feynman, and like him other physicists, thought that the field of computing focused on mathematical horizons and moved too far away from nature, and that physics could be more connected to the world of information," explains Gil. "In a historic lecture he gave in 1981, Feynman claimed that there was nothing to give a classical computer to deal with particle simulation, because nature is not classical. He said, 'If we want to simulate nature, we need a machine that behaves like nature, in a quantum way.'" In 1998, this vision was realized, when the first quantum computer was built at the University of Oxford in Great Britain.
A quantum computer utilizes the enigmatic properties of quantum theory, those that are not fully understood by us, to perform calculation operations. In a normal computer, the basic unit of information is a "bit", which can have one of two values, 0 or 1; Using such bits makes it possible to perform any calculation imaginable - although some of these calculations may take a very long time. In a quantum computer, the qubit, thanks to superposition, represents not one absolute value, but a distribution of values. "You can think of it as a question of more dimensions: one and zero are just the ends, the poles of a coin for example, but it can also have a sideways tilt," explains Gil. Using statistical approaches it is possible to examine the state of the qubit and obtain useful results. This probabilistic approach is not suitable for every problem, but in solving certain problems it is infinitely more efficient than the classical computer's search for an absolute answer.
"Because of the entanglement effect, it is also possible to cause the qubits to influence each other," says Gil. And since each qubit represents an entire field of possibilities, each addition of a qubit increases the number of possible connections between the qubits with exponentially increasing power (in the classical computer, on the other hand, the addition of bits grows linearly). At the moment, IBM holds the record for qubits: last year it unveiled a quantum processor with 127 qubits, and its stated goal is to launch a processor with 433 qubits this year, and a processor with 1,021 qubits next year.
Three degrees colder than outer space
This ambition is more pretentious than it seems. It turns out that "building a machine that will behave like nature" is a complex story like no other: the qubits are very sensitive to outside influences, which makes building a computer a very complicated and expensive business. "The quantum computer is very powerful, but at the same time also very delicate," explains Gil: "It utilizes physical processes that occur in the world, but such processes are a system in which everything is connected, everything affects everything, and this can disrupt the results: if energy from the outside world goes inside and connect to the qubits, this will make them behave like normal bits, and thus the unique ability of quantum computation will be lost. Therefore, a quantum computer must be very isolated from the entire environment. The big challenge is to produce a system that is sufficiently isolated from the outside world, but not too isolated."
When I try to find out what the cost of building a quantum computer is - and IBM has already built 40 of them - Gil avoids a clear answer, but it is enough to hear what this effort entails: "There are several different approaches to building a quantum computer; IBM chose a cryogenic approach, meaning deep freezing, and the use of superconductors. The temperature in the computer is close to absolute zero: at the bottom of its case the temperature is minus 273 degrees Celsius—three degrees less than the temperature of outer space, and less than one degree above absolute zero. The temperature should be close to absolute zero, but not reach it, because then there is no movement at all, Not even of the atoms."
The result is a cooling and protection case that resembles a water heater in its shape, and inside it has the calculation unit, whose shape gave it the nickname "chandelier" according to Gil and his team. "Inside the layers of protection there is a cylinder with the processor in it. Even if only a fraction of an energy particle enters the computer, literally a fraction of nothing, it will be enough to disrupt the results," Gil clarifies.
The great sensitivity, and the protection requirements derived from it, mean that the quantum computer is quite cumbersome: in the newest models, which try to include more and more qubits, the case already reaches a height of several meters. To some extent it is reminiscent of the first generations of classic computers, which looked like huge cabinets. Those classic computers kept getting smaller and smaller, until today we squeeze millions of times more computing power into a simple smartphone, but in the case of quantum computers, we cannot expect a similar process: "The quantum computer requires unique conditions that cannot be produced in a simple terminal device, and this will not change in the foreseeable future," Gil explains. "I believe that quantum computing will be a service that we can access remotely, as we access cloud services today. It will work similar to what IBM already enables today: the computer sits with us, and we make it possible to access the 'brain' and receive answers. Of the 40 computers we have built since 2016, today 20 are available to the public. About half a million users all over the world have already made use of the capabilities of the quantum computer we built, and based on this use, about a thousand scientific publications have already been published."
Google and Microsoft are heating up the competition
IBM is not the only company participating in the quantum computing race, but Gil exudes full confidence in its ability to lead it: according to him, most competitors only have parts of the overall system, but not a complete computer available to solve problems. Google, as mentioned, is a strong contender in this race, and it also allows remote access to its quantum computing service, Google Quantum AI; Microsoft is also working to provide a similar service on its cloud platform, Azure.
Meanwhile, quantum computing is a promise "on paper". The theoretical foundations for this revolution were laid already 40 years ago, the first proofs were presented more than 20 years ago, the industry has been buzzing around this field for several years - and we still haven't seen uses that would serve a regular person.
"If you go back to the 1940s, when the first computers were invented, you will see that even then the uses and advantages of the new invention were not clear. Those who saw the first computers said, 'Oh, great, you can use it to crack the code of encryption machines in wars, maybe even calculate routes of ballistic missiles, and that's it. Who's going to use it? Nobody,'" Gil laughs. "In the same way, the success of quantum computing will depend on its uses: how easy it will be to program, how large the community of users will be, what talents will get there. The quantum revolution will be led by a community, which is why education for this field is so important: we need more and more smart people to start to think 'how can I use quantum computing to advance my field'.
"What is beginning these days is the democratization phase of quantum computing, which will allow anyone to communicate with the computer without being an advanced programmer in the field: it will be possible to approach it with a question or a task that will be written in the classical languages of one or zero. That is why we are already seeing more use of quantum computing capacity today.
"There are also many startups that do not actually work to establish a quantum computer, but focus on various components of this world (for example, the Israeli company Quantum Machines, which develops hardware and software systems for quantum computers, and last July was selected by the Innovation Authority to establish the Israeli Quantum Computing Center). The activity of such companies creates a completely new ecosystem, thus promoting the industry and accelerating its development, just as is happening today in the field of ordinary computers. IBM will not rely only on itself either: we would like to benefit from the innovation of smart people in this field, of course also in Israel.
"I am convinced that the big bang of quantum computing will happen in this decade. Our ambition at IBM is to demonstrate 'quantum supremacy' already in the next three years. I believe that the combination of advances in artificial intelligence, together with quantum computing, will bring about a revolution in the industry of the kind that Nvidia made in its market (Nvidia developed unique processors for gaming computers, which made it the chip company that reached a billion dollar revenue the fastest.) Quantum computing can generate enormous value in the industry. It is phenomenally difficult, but it is clear to me that we will see the uses already in the current decade."
The Nobel Prize opens a new horizon for quantum computing
Quantum computing has ignited the imagination of researchers for many decades, but until now it has not left the confines of laboratories. However, the awarding of the Nobel Prize to three researchers in the field indicates that the vision is becoming a real revolution. Alain Aspect of France, the American John Clauser and Austrian Anton Zeilinger received the award for research they conducted (separately) since the 1970s, in which they examined the phenomenon of quantum entanglement (described in the article), proved its existence and laid tracks for its technological use.
The awarding of the Nobel Prize to the entanglement researchers proves that quantum computing is more than a mental exercise for a sect of physicists, and is a defining moment for companies that invest capital in the development of the field. They are pushed to this effort due to a fundamental change in the world in which they operate: in recent decades, the world of computing has operated according to "Moore's Law", which foresees that the density of transistors in computer processors will double every two years in a way that will increase the computing power of these chips. However, as the industry approaches the physical limit after which it will be impossible to cram more transistors onto a chip, the need to develop a quantum computer has become acute.
The numbers also signal that something is happening in the field. In 2020, the scope of the quantum computing market was less than half a billion dollars, but at the end of 2021, in a signal that the vision is beginning to be realized, the research company IDC published an estimate according to which in 2027 the scope of the market will reach $8.6 billion and investments in the field will amount to $16 billion (compared to $700 million in 2020 and $1.4 billion in 2021). IBM CEO Arvind Krishna also recently estimated that in 2027 quantum computing will become a real commercial industry.
First published: 09:00, 17.10.22
