Quantum computers are touted to be the next best thing the world of technology will offer. IBM, Microsoft and Google have been in competition to set up a quantum computer for real. In September 2019, Google claimed to have achieved ‘quantum supremacy’—the first time a quantum computer has outperformed a traditional one. An article published in the England-based Financial Times newspaper stated that researchers at Google claim to have solved a really difficult problem in seconds with the help of a quantum computer—which a supercomputer could not do. The research paper, however, was yet to be formally vetted by peers in the field. It only became public after having appeared briefly on the National Aeronautics and Space Administration (NASA) website — apparently, some of its researchers were involved in the project.
The research paper mentions that Google used Sycamore, a 53-qubit processor, to generate a sequence of millions of numbers. These numbers appeared as if they had been randomly generated, but they conformed to an algorithm generated by Google. A supercomputer checked some of these values and they were correct. The quantum computer has claimed ‘supremacy’ because it reportedly did the task in 200 seconds—it would have apparently taken a supercomputer 10,000 years to do!
Soon after, a team of scientists from IBM, in a paper published on the physics website arXiv and in a blog entry at the research website of the company, challenged the Google claim: the calculation would not take 10,000 years on a regular computer but could be accomplished in just two-and-a-half days!
Developments
The ideas governing quantum computers have been around since the 1990s but actual machines have come to exist only from 2011.
In 1994, Peter Shor, when working in Bell Labs, came up with an algorithm that a quantum computer could use to factor big numbers. The intention was to break cybersecurity codes now in common use.
In 2012, theoretical physicist John Preskill of the California Institute of Technology coined the term ‘quantum supremacy’, to refer to the point at which a quantum computer can outperform a classical or traditional computer. It refers to a quantum computer being able to solve a problem that a classical computer cannot.
IBM, Google, Microsoft and D-Wave Systems have been involved in some way in designing and building starter versions of quantum computers and putting them online.
IBM has had a set of starter computers called IBM Q Experience available online since 2017 where anyone can log on and write and run programs. At least some 130,000 people had used it by 2020, running 17 million experiments and publishing some 200 papers, according to the company. It offered a tour of its quantum operation which involved showcasing a display quantum computer: a series of the gold-coloured platform hanging one from another, with chips, wires, capsules and curled silver tubes. The real working computer, called IBM Q System One, with only 20 cubits, was encased in a 9-foot-wide cube of black glass sealed in an extremely cold atmosphere where noise and interference were blocked. System One went online in January 2019.
But what are quantum computers and how are they superior to regular computers and supercomputers? Importantly, what is quantum computing?
About Quantum Computers
Quantum computers exploit the principles of quantum mechanics, so they can easily tackle computational problems that may be tough for the computer of today. With quantum computers, the size of the numbers and number of inputs involved grows bigger. There are only very few quantum computers that have been set up. Unlike the desktops or laptops, they resemble the air-conditioned server rooms of many offices or the stacks of central processing units from desktops of yore that are connected by tangled wires and heaped in freezing rooms.
Conventional computers process information in ‘bits’ or 1s and 0s, i.e. they can process a ‘1’ or a ‘0’ at a time. Every app, website visited, photograph taken is made up of millions of these bits (in combinations of 1s and 0s). The world’s most powerful supercomputer can juggle 148,000 trillion operations in a second and requires about 9000 IBM CPUs connected in a particular combination to achieve this feat. But when we see nature, we find that things are complicated. We see complexity and so we find that information that cannot be processed as ‘on’ and ‘off’. To actually simulate the things in nature, scientists need a better way of handling calculations that can address what they see as ‘uncertainties’. Hence the importance of quantum computers.
Quantum computers compute in what are called ‘qubits’ (quantum bits), which can be in ‘superposition’ rather than just on or off: they are both on and off at the same time or somewhere on a spectrum between the two. They exploit the properties of quantum mechanics—the science that governs how matter behaves on the atomic scale. As stated, in this scheme of things, processors can be a 1 and a 0 simultaneously, a state that is called quantum superposition. In quantum computing, one can use information which would involve complex calculations. These calculations stringing together multiple qubits can help solve problems that present-day computers may find it impossible to solve.
Quantum computers speed up computation: a machine with less than a 100 qubits can solve problems with a lot of data that are even theoretically beyond the capabilities of the most powerful supercomputers. Because of quantum superposition, a quantum computer — if it can work as planned — can mimic several classical computers working in parallel.
Advantages
The quantum computers do not just do things faster and more efficiently but they can do calculations that even the best supercomputer cannot do today. It is speedy, as it is able to simulate several classical computers working in parallel.
The speed and capability of supercomputers are restricted by energy requirements, even as they need more physical space. Processing huge amounts of data quickly is a real-world problem that can be tackled faster by quantum computers. So, for instance, if the government wanted to scrutinise particular persons from a database of a million profiles, a normal computer would have to scan each one of those profiles which would amount to a million steps. But with a quantum computer this would be done faster—say, with 1000 steps instead of a million—which would mean reduced processors and less of energy requirement, time consumed, and so on.
Quantum computers can have a variety of applications. They can be used anywhere where a complicated system that one is uncertain about, needs to be simulated and worked out. This could involve predicting the financial markets, forecasting weather and climate, modelling the behaviour of electrons or understanding quantum physics.
Encryption Systems Cryptography is seen as an area of key application. Encryption systems used by the government for various purposes and various organisations to secure data flowing through the internet rely on the difficulty of breaking down large numbers into prime numbers (called ‘factoring’). Classical computers cannot break encrypted data easily: it is time-consuming and expensive. But quantum computers can do it easily. Encryption systems are used by governments, their military, intelligence, financial and administrative departments, to secure data. It is also used by public institutions, financial organisations, companies and websites to secure data. With quantum computers coming on, the encrypted data can be broken down. (The solution to this is ‘quantum encryption’: quantum encryption keys cannot be copied or hacked easily).
Limitations
The latest Google experiment does not imply that their quantum computer can solve every challenging problem. In fact, the number-generating task (generating a sequence of millions of numbers in 200 seconds, as using Sycamore did) was the equivalent of having a Ferrari and a truck compete in a race and, on the car’s predictable victory, declare that the Ferrari could do everything that a truck could do, which is absurd. IBM and a few other private establishments also have quantum computer prototypes. But a common ailment of them all is that they have their own unique propensity to errors; they are less amenable to executing real-world problems compared to supercomputers. Though the company D-Wave Systems claims that automobile-maker Volkswagen used its quantum computers to figure out how best to control a fleet of taxis in Beijing relying on data from 10,000 cars, the research paper describing this experiment does not say how the proposed solution is better than algorithms that are currently used to optimise traffic flow.
Quantum computers have been theorised about for many years but it is only recently that they are making news. Why? Because these computers have a lot of limitations.
A quantum chip can never find a place in the smartphone or the laptop because they are extremely sensitive to interference. Any slight electrical interference can knock a qubit out of its delicate state of superposition. So these computers have to be isolated from all electrical interference and chilled down to close to absolute zero temperature (colder than outer space).
At present, the best quantum computers have about 50 cubits, which makes them powerful because every qubit added means an exponential increase in the processing capacity. However, as they have high error rates as well because of the problems with interference, the computers are not reliable. There is prone to errors also means that while they may be good at predicting probabilities, they would do a lot of harm if used to compute annual financial statements of a company.
The big breakthroughs reached, as with Google stating it had achieved ‘quantum supremacy’, have been done in a controlled setting. Or, the problems used are those for which we already know the answer (by using ordinary computers). So reaching quantum supremacy as claimed does not mean quantum computers are actually ready to do anything useful. Though researchers have done work developing the algorithms that quantum computers would likely use, the computers themselves have to be developed much more.
Breaking down Encryption Several encryption systems used in government surveillance, security and intelligence, and military, and security applications as well as by financial institutions, academics and businesses are premised on computers being unable to handle mathematical problems that are computationally demanding beyond a limit. It is feared that the use of quantum computing and its ability to break encryption codes will impact their use of encryption systems. For quantum computers, in theory, can surpass those limits. However, that’s only in theory! Actually breaking banking grade encryption is another thing.
The ultimate goal of quantum supremacy is to use qubits to crack encryption codes. Current encryption standards may require a quantum computer to have several thousand logical qubits working in tandem perfectly in order to break encryption. This requires millions of qubits of the kind that powers the quantum computer Sycamore to make ‘logical qubits’. The 53 at Sycamore’s disposal do not mean anything. And each of those several thousand logical cubits working in tandem will need to be protected by many hundreds more—to protect against errors introduced by outside noise and interference. However, there are other approaches to designing quantum computers and with it there may be cleverer ways to solve problems using them. If technological breakthroughs came close to posing a real threat to governments’ use of encryption systems, they may well take to set up more advanced quantum computers that can prevent the threat.
Moreover, new mathematical methods or techniques may come that would allow the development of ordinary computers so that they execute the same task faster. In which case, there would not be any need for quantum computers!
What the experiment has shown is that quantum computers can be used for a real-world task. Private entrepreneurs and academics would be keen to invest time and money to improve them and customise them to real-world problems. D-Wave Systems, which already has a 1000-qubit system at NASA, is ready to commercially launch a 5000-qubit system.
In all, quantum computing is getting on, even if Google’s claims of using its quantum computer with a few errors to make it outperform today’s supercomputer, cannot be rightfully called ‘achieving quantum supremacy’. However, quantum computers would remain out of bounds for commercial applications in the near future. And even if it enters the commercial arena, it would only complement what the regular computers do, and not compete with them.
|
In India There are no quantum computers in India. The government has announced 2 initiatives in this field: a networked programme on Quantum-Enabled Science & Technology (QuEST), launched by the Department of Science & Technology in 2018, and the National Mission on Quantum Technologies & Applications (NMQTA). The Department of Science and Technology (DST) announced that it was committed to investing ₹80 crores over a period of 3 years to accelerate research for the ostensible plan of having a quantum computer built in India within the next decade. |