10.3.22

The Hope and Risk of Quantum Computing

Category:
Blog Post

The Hope

 

For years, we have been told that quantum computers are coming and that they are going to change everything. Today, they are not just the future. Quantum computers are already a reality and more powerful generations of quantum chips are produced every few months. At least half a dozen companies build and sell quantum computers and a thriving industry has grown up around software design, the manufacture of components, and application planning. Quantum is already here.

 

Quantum computers have been around for long enough that we can see that they will be capable of much more in the future. Just as Moore’s Law made it clear that the hulking classical computers of the 1960’s would lead to more powerful systems, the pace and scale of the quantum industry demonstrate that quantum computers will reach a level of sophistication which will allow them to solve problems that no classical computer is capable of.

 

The advantages of quantum computing will not allow it to replace classical computers for every computing task. Quantum computers are very good at modeling a complex system and finding optimal solutions from among a large set of variables that would overwhelm classical computers. These are not traits that we need in word processors or web browsers. But they are traits that are important in fields including chemistry research, financial analysis and cryptography.

 

Classical computers work on the basis of ones and zeros. A transistor is either on or off. In quantum computing, systems take advantage of the different set of rules for physics at the atomic and subatomic level. Quantum particles with a trait such as the direction of their spin can be spinning in one of two directions (‘spin’ is a simplification of complex physics, but fine for understanding the basics), that can be utilized for computation in a way similar to classical transistors having a state of one or zero. But at the quantum level, when information about a particle does not escape, it is simultaneously spinning in both directions. It exists in a state of multiple probabilities as to its spin. Only by observing what it is doing does the particle settle into a definite spin. This is a rough example of what physicists call “superposition,” one part of the system of quantum mechanics that describes the different rules by which very small things work.

 

The basic units of quantum information are called qubits. Anything small enough to exhibit the effects of quantum mechanics could probably be used somehow as a qubit. Individual atoms, molecules, simple lattices of molecules. But to be a good candidate for quantum computing, the chosen material needs to be easy to manipulate and isolate from outside interference. It also has to be something that can be scaled up to build computers capable of doing meaningful work. Different quantum computing companies and research labs have chosen a range of quantum architectures, from chains of individual ions to tiny superconducting loops of metal.

 

With their options for one, zero and a superposition of both, qubits have a distinct advantage over classical transistors. Quantum computers are able to represent every possible permutation of available qubits in superposition, so while a classical computer with 4 bits can be in any one of 2^4 states the quantum computer has 2^4 states in superposition with one another. By successively applying quantum operations on the qubits, the quantum computer can make the correct answer the most likely solution to be measured out of this astoundingly large search space. By running the problem repeatedly, confidence in the correct solution increases.

 

An example that illustrates the usefulness of this is what is known as the traveling salesman problem. A traveling salesman wants to visit a list of different cities on a map without repeating any of them and would like to know the most efficient route among them. Classical computers start having a lot of trouble with this at around fifty cities. There is no supercomputer on Earth that can do it perfectly with a hundred cities. That is because the classical computers that try to solve the problem by running through each possible route in sequence will take thousands of years to model them all. Even with some computational tricks more sophisticated than checking every route the classical computer will eventually be beaten, because every time you add a city you double the number of possible routes.

 

Quantum computers have strong potential for working on traveling salesman problems because they don’t need to go through each route consecutively. They look at the entire network of cities and define how an optimal solution would look, subject to the limit imposed by the number of qubits in the system. Quantum computers have opened up new possibilities for optimal solutions to traveling salesman problems, which are related to multiple classes of optimization problems that many quantum researchers and developers are working on.

 

Already, useful applications have been demonstrated for this technology. how their quantum processors can be used for optimization in multiple areas of finance. IonQ has a partnership with Hyundai to use their quantum computers to model the characteristics of potential new chemical compounds to use in battery development.

 

As quantum computers scale up, there is clear potential for this technology to make the world a better place. More efficient batteries in electric vehicles, faster logistics for shipping and delivery, less risk for investors’ portfolios. Quantum computers could have just as profound an effect on society as classical computers have.

 

In 2016, IBM first tested their first five-qubit Canary system. Several generations of processors later, they now provide customers with the 127 qubit Eagle. A 433 qubit Osprey processor will debut before the end of 2022, with the 1,000 qubit Condor to follow in 2023. Each additional qubit in a quantum computer doubles its computational power. This technology is scaling up at an astonishing pace.

 

Quantum technology is also helping to improve security. For example, Oak Ridge National Laboratory developed a device that monitors the quantum characteristics of large numbers of photons in order to generate truly random numbers. Generating random numbers is a major challenge in cryptography. A tally of, say, the number of bytes in every file on your laptop might seem random at first but as it grew it would eventually show a pattern that a computer could model and then predict. Most sources of seemingly random numbers are susceptible to this. The quantum randomness of large numbers of photons can provide a steady source of truly random numbers.

 

Qrypt has exclusively licensed this technology from Oak Ridge National Laboratory and in addition to other sources in its advanced technology pipeline, incorporates it into the quantum random number generators that help to secure our products.

 

The Risk

 

Among the many positive expected applications for quantum computers, there is a major problem  that is most concerning. That ability which we have described for a quantum computer can equally be applied to kill most types of encryption currently used.

 

RSA, the most common form of encryption used today, depends largely on the fact that it would take thousands of years to try every possible solution to decrypt a given file. Such a “brute force attack” doesn’t work well using classical computers because they have to try solutions one at a time (unless the file was encrypted using numbers which were not as random as they needed to be). As quantum computers become more powerful, they will increasingly get closer to being able to rapidly crack standard encryption.

 

The tool for this has already been developed. Shor’s algorithm. Widely understood and accepted, it is a mathematical method for identifying the prime numbers of an integer. It requires a type of quantum computer that’s not available yet, but as systems with more, higher quality qubits become available, Shor’s algorithm will crack RSA.

 

In recent years, some organizations have transitioned away from RSA in favor of elliptic-curve cryptography. According to the NSA, elliptic-curve is also vulnerable to Shor’s algorithm and the Agency expects that quantum computing will render it useless even before it cracks RSA.

 

Your encryption key is like a needle in a haystack. Classical computers have to pick through each piece of hay. Quantum computers will be more like a magnet that pulls the needle right out.

 

Nobody knows exactly when this will happen, but the quantum naysayers who insisted that quantum computers could never work at all have already been proven wrong by the flourishing quantum industry and range of quantum computers currently in use by governments, universities and corporations around the world. You can even access quantum computers today using major cloud services Amazon Braket , IBM Cloud, or Microsoft’s Azure. As major quantum computing companies develop new generations of quantum computers, they often make them available on the cloud to customers all around the world.

 

Is the danger in two years? Five years? Ten years? Will it even be publicly known when RSA or elliptical-curve encryption is first cracked by a quantum computer, or will the victims of the hack simply be scratching their heads when private data is suddenly being leveraged by a hostile government? Nobody has a definite answer to these questions. There are no vetting processes used by quantum cloud providers to determine what customers are doing on those commercial quantum computers.

 

There is plenty of encrypted data which is poorly protected today because the encryption is supposedly secure enough that it doesn’t matter if hackers and nation-states get it. They are capturing the encrypted data as it travels through the internet because a lot of that data will still be valuable years from now, once more powerful quantum computers are available. Hackers can harvest encrypted data today, wait for the technology to improve, and then decrypt it. They are capturing encrypted social security numbers and banking info, medical records, DNA data, keys to cryptowallets, and industrial trade secrets. Most of this will still be valuable years from now.

 

The scientific journal, Nature, wrote in February of 2022 that “…the machines that don’t yet exist endanger not only our future communications, but also our current and past ones. Data thieves who eavesdrop on Internet traffic could already be accumulating encrypted data, which they could unlock once quantum computers become available, potentially viewing everything from our medical histories to our old banking records.”

 

If quantum computing will ever be a threat to encryption, it is already a threat today.

 

There is also the internet of things to be concerned about. Many modern automobiles, medical technology, farm equipment, water treatment plants and power grids have components which are remotely accessed through quantum-vulnerable cryptography. The risk is present in industries that may not think of themselves as being in the business of securing data, and these types of systems get put in place for decades.

 

There are two actions which your organization needs to take in response to the looming quantum threat to cryptography. The most obvious one is that everything needs to be migrated to quantum-safe standards over the next few years before really powerful quantum computers arrive. But the more pressing action is to convert your sensitive data into a quantum-secure form immediately.  Don’t encrypt your private data in a form that will ever be vulnerable to quantum computers.

 

Qrypt offers encryption that does not just depend on being computationally very difficult for a quantum computer to crack. We provide encryption which is mathematically proven to be impossible for a quantum computer to break. We do this using truly random, quantum-derived numbers, without the need for keys to even be transmitted. To find out how Qrypt can help your organization contact us at info@qrypt.com or sign up for our monthly newsletter.