Quantum Computing Use Cases - Time to Start Dreaming!

April 8, 2022
Quantum Computing Use Cases - Time to Start Dreaming!

The development of the quantum computer could be as elemental of a change in IT as the creation of the transistor. Quantum computing holds the prospect of solving problems that were previously beyond our capabilities or at best extremely time consuming to solve using classical IT methods. While there are many connections to the world of classical computing, quantum computing should be seen as a different mode of computing, and as a mode in its early days at that.

How we can use this new mode is still being discovered. Some use cases, at least in terms of fields, can be seen already. At Proxet, given our fascination with all things AI/ML and our extensive experience in industries such as finance and medicine, the promise of quantum computing is enticing. In this report, we will cover the basics of quantum computing including its strengths and drawbacks, how it relates to the fields in which use cases are envisioned, and possible applications of quantum computing. In particular, we’ll focus on the interconnect between quantum computing and AI/ML.

What Is A Quantum Computer?

Let’s step back to Square One for a moment and look at how quantum computers work, what they do better than classical computing, and how they might interface with your desktop or notebook in the future, or if you want, interface with your classical computer before you get done reading this article.

Classical computing uses two states, called 1 and 0, and from this binary form, has been able to create computing as we know it. Also, whether processing multiple bits in series or in parallel, or in packets, each individual bit of information is essentially independent. It may link with other bits or packets formally, but essentially, you could pluck it out of the data stream and there would be no essential change other than an error that might cause the packet to be sent again.

Sidebar: Quantum coprocessor in your PC?

Yesterday, the math coprocessor, today, the graphics card. And tomorrow – the quantum adapter? In the 1980s, the mighty 80287 math coprocessor was the must-have addition for serious desktop computing. These days, solid state drives and especially graphics cards with their own memory are marks of prosumer or higher-end setups. Tomorrow? Maybe quantum coprocessors for those who don’t want QCaaS.
Don’t look for one on your favorite e-retailer just yet, however. The environment quantum computing requires includes a vibration-free setting, for example. The size of the installation and the cryogenic environment are expected to remain prohibitive for the time being.
CHINESE $5000 machine – but where is it?

Quantum computing is completely different at its core. Through one of several methods, the bit is held in what’s called superposition, which allows it to be both 1 and 0, as well as in other combinations, all at the same time. As a bit in a quantum state, it gets called a quantum bit, or qubit. For the qubit to remain in a superposed state, it needs to not be interacted upon by its environment. If it does, it decomposes into either a 0, or low energy state, or a 1, or high energy state and the quantum intermediate states are lost. Maintaining superposition is one of the key challenges of quantum computing, and is usually done at temperatures close to absolute zero.

Image by Proxet, Classical Computing and Quantum Computing
Classical Computing and Quantum Computing

Furthermore, instead of being a discrete unit, qubits connect to each other at the quantum level in what’s called entanglement. Entangled qubits share the same state for as long as they remain at the quantum level. As we shall see, entanglement can, under some circumstances, be useful, and always should be under control. This is because every time a qubit is added to an entanglement, the connections don’t add arithmetically, they add exponentially. The amount of computing power becomes tremendous very quickly. When IBM announced the debut of its Eagle 127-qubit Eagle QPU, the press release stated that, “In fact, the number of classical bits necessary to represent a state on the 127-qubit processor exceeds the total number of atoms in the more than 7.5 billion people alive today.” And as ZDnet notes, at 275 qubits, “we can compute with more pieces of information than there are atoms in the observable universe.”

Introducing The Quantum Computer

But, how do you work with qubits?

Let’s take a block-diagram level look at a quantum computing setup.

Image by Proxet, Quantum Computing Components
Quantum Computing Components

Classical computing uses registers and logic gates as the fundamentals of hardware, and quantum computing works with quantum registers and reversible gates. Despite the similarities in nomenclature, though, there are more differences than similarities. Moreover, unlike classical computing, there is more than one way to work with a qubit. Let’s look at them for a moment.


Working with photons has some convenient advantages like operating at room temperature. Photon-based computers can plug into fiber-optic telecoms infrastructure, which is already used extensively in networking. These advantages stem from the weak interactions of photons with their environment, even at room temperature.

These advantages are offset by issues with fault tolerance and error correction.  Companies working on quantum computing technology based on photonics include Amazon Quantum Solutions Lab, PsiQuantum, and Xanadu.

Trapped ions

Honeywell and IonQ are two of the companies working on using microwave or optical signals sent through free space or waveguides to the qubits. The qubits themselves sit in chains of static ions in potential wells. However, it is difficult to isolate a single individual ion’s motions as the number of ions in the well increases. Measurement is also difficult with this method.

Semiconducting material

One promising route for quantum computing is the use of semiconductors. If a microwave or magnetic field is applied to selenium or germanium or diamonds, then superposition, entanglement, and other quantum properties can be seen. One important advancement is that this can be done at room temperature.

Superconducting material

Superconducting qubit systems were the first successful method for contemporary quantum computing. These systems send microwave and low-frequency electrical signals through wires running into refrigerators to reach the qubits inside the controlled environment.

Google and IBM are among the leaders in this direction of quantum computing. Given the resources of these two tech giants, it comes as no surprise that extensive research is going on regarding the various properties of superconducting materials for quantum processors.

There is a lot to be done. Shrinking the size of bits in classical computing has played an essential role in the creation of smaller machines, and while the transistors of classical computing are down to the single-digit nano scale, superconducting qubits are comparative monstrosities millimeters in size.

Being able to affect even a component comprising a qubit can have surprisingly large effects on multiple factors at the same time. For example, in January, 2022, MIT researchers announced that they had found a way to reduce the size of the material used to build capacitors within qubits by factors of 10. Doing so would not only make qubits smaller, given the smaller size of the capacitors, but would also reduce the amount of crosstalk, or internally generated interference due to energy leakage between qubits. While quantum entanglement is a good thing, crosstalk, quantum or otherwise, is not.

Quantum Processing Unit

A quantum processing unit (QPU) is a small device. Depending on the type of chip, it could be smaller than a DB-25 connector. A QPU includes QRAM (with register and gates) and a control unit, which is still small. The rather large size of a quantum computer comes from the need to refrigerate the QPU to absolute zero. Classical controller interfaces define the interaction between the host CPU and the QPU.

Quantum Registers

Unlike classical registers, which work with a number of flip-flops, quantum registers only count the number of qubits. A quantum register holds all possible input data configurations simultaneously. In this manner, quantum registers differ greatly from classical ones.

An example can quickly show what is meant. In classical computing, a register two bits in size  holds states of 1 or 0, with four possible outcomes (0,0; 0,1; 1,0; 1,1). A quantum register two bits wide would simultaneously hold all of the possible states for the two qubits.

Quantum Reversible Gates

Reversible gates can have their input reconstructed based on the output. This is important, because logical reversibility allows for:

  • Reversing quantum circuits, which shows that the output could have been derived from only one input.
  • Greater energy efficiency, as computation does consume energy.

Using A Quantum Computer

Quantum Computing as a Service (QCaaS) aside, users usually plug into a quantum computer through a classical computer called a host processor. This processor connects to the QCU through a high bandwidth connection. Conventional operating systems are utilized on the host processor.

Quantum Hardware Challenges

The hardware that is specific to quantum computing is now out of the lab, but is still maturing. How to handle the cold of superconducting materials is just one of the issues manufacturers face. Others include:

Furthermore, instead of being a discrete unit, qubits connect to each other at the quantum level in what’s called entanglement. Entangled qubits share the same state for as long as they remain at the quantum level. As we shall see, entanglement can, under some circumstances, be useful, and always should be under control. This is because every time a qubit is added to an entanglement, the connections don’t add arithmetically, they add exponentially. The amount of computing power becomes tremendous very quickly. When IBM announced the debut of its Eagle 127-qubit Eagle QPU, the press release stated that, “In fact, the number of classical bits necessary to represent a state on the 127-qubit processor exceeds the total number of atoms in the more than 7.5 billion people alive today.” And as ZDnet notes, at 275 qubits, “we can compute with more pieces of information than there are atoms in the observable universe.”

  • Isolation

Temperature isolation is the greatest issue for quantum computers using superconducting materials. Quantum coherence is greatly affected by heat and light, and the loss of coherence   kills superposition and entanglement along with the information the qubit stores. Typically superconducting quantum computers are kept at approximately 0 Kelvin.

  • Signal control

Work with a qubit requires rotating it by means of a quantum logic gate. These rotations are not always accurate, and there can be consequences. For instance, if an algorithm calls for rotating a qubit by 90 degrees but it rotates by 90.1 degrees, it could result in an incorrect output.

  • Quantum error correction

Quantum noise such as decoherence is taken into account by quantum error correction (QEC). Classical computing manages error correction through redundancy. Bits used to encode information are copied and stored many times in order that they can be checked as to whether they are the same. If there are changes to some of the copies, then the information that is most prevalent is considered to be the true version.

However, quantum computing operates under the no-cloning theory. This theory holds that you cannot perfectly copy an arbitrary quantum state. Another issue for quantum error correction is wavefunction collapse. Classical computing allows for arbitrary properties to be measured without disturbing the content that is encoded. For quantum computing, error correction by means of the measuring of qubits can lead to the wavefunction collapsing and erasing the encoded information.

What exactly do ç look like? The image that usually accompanies popular articles shows a myriad of wires like a biplane and metal architecture, as if a Cray 1’s internals were dipped in gold and left to hang upside down. The comparison to an old supercomputer is apt, given that quantum computing is in early days, and is a maturing technology just as aviation and supercomputing were at one time.

Image by Proxet, Quantum Computers
Quantum Computers
Image by Proxet, Quantum Computers
Quantum Computers

But these chandelier-style fabrications are not the processor itself. Quantum processors like the IBM Quantum Hummingbird do not look all that different from most other chips at first glance.

The early days of quantum computing have also given rise to different answers to questions such as stability and error correction. A single architecture may come to dominate completely over time, or like biplanes that are still used for specific use cases, some of the others might survive because there is a need that a particular architecture fits best. Currently, gate model architectures, such as those used by IBM and Google, and D-Wave’s quantum annealing are the two main routes for building a quantum computer.

Regardless of the architecture, there are issues common to them all. Examples of the problems that quantum computer designers and builders face include:

  • Physical requirements such as temperature;
  • Qubit creation;
  • Fault tolerance

Physical Requirements

When quantum computing first came to the attention of mass media, one salient point that caught the public’s eye was the environment required for the computers to work. Of the various ways to create a qubit, the most exotic route yet easiest to understand was the use of extreme cold. With this route, qubits are created at 20 milliKelvin, or just above absolute zero. This route is unlikely to make it to mobile or desk-top applications in the foreseeable future. It is the most common method for creating qubits commercially, though, despite knock-on issues such as the interface between the QPU and the rest of the system. The need for hardware that can operate at or near the QPU’s operating temperature has created a race no less important than the number of qubits in the processors, even if it has not received the same amount of press attention.

The quantum cryogenic ecosystem that is developing now includes QPU controllers from both Google and Intel. The Intel chip is designed to be as close to the QPU as possible, according to ZDnet, while ieee.org notes that the Google chip is intended for operation at a balmy 4 Kelvin, which is the temperature found about 50 cm away from the QPU. Low Noise Amplifiers (LNAs) and multiplexers that can eliminate some of the wiring in such an environment are also in the cards.

Qubit Creation

When the first classical computers were built, vacuum tubes were the only way to create the electrical and logical structures that were a “computer”. These gave way to the transistor and the integrated circuit, and eventually, to virtual machines. Quantum computing has not followed such a neat development path; the industry is closer to that of the automobile, with internal combustion (hydrocarbon and steam) and electric motors providing propulsion.

In quantum computing, ultra-cold devices are at a stage similar to the internal combustion engine with petrol and diesel fuels were at shortly before mass production began. Google and most other companies working with cryogenic quantum machines use a quantum logic gate. D-Wave, which recently installed the largest quantum computer in Europe, uses a process called quantum annealing (QA). The QA model offers several advantages in terms of operations, control, and environment, and the D-Wave machines already contain thousands of Qubits. However, the quantum annealing method is best suited for optimization problems, and does not have the same range of operations as quantum gate machines.

Another way of creating and manipulating a qubit is to trap it in a magnetic field and use a laser to interact with it. This is the approach used by IonQ, which became the first full-stack quantum computing startup to go to the New York Stock Exchange.


Given the rarity of quantum computers, remote access is by default the simplest way to make them accessible to a broader range of people, especially in an enterprise setting. Quantum Computing as a Service (QCaaS) already exists as an outgrowth of this need. McKinsey sees QCaaS as most people’s way of accessing quantum computing for the foreseeable future, and the market is likely to explode. Quantum Insider sees the QCaaS market rising from an estimated less than USD 50 million in 2020 to USD 4 billion by 2025 and USD 26 billion by 2030, according to an August 2021 report.

Fault Tolerance

While the earliest days of quantum computing are past and commercial machines are available, these are similar to the barnstorming days of early flight. Daring feats are being done on machines that aren’t quite yet perfected. In the case of quantum computing, quantum superiority may have been achieved, but for accuracy, overall, a classical computer is required. Creating a quantum computer that can reach the level of error correction needed to ensure the accuracy of results for the pharmaceutical industry, for example, may still be years away. However, companies are currently working on ways to improve both the quality of the qubits and processing as well as creating hybrid systems that connect the quantum computer to a classical one. For example, D-Wave installed a Leap series quantum annealer at the Forschungszentrum Jülich Supercomputing Centre in Germany in January 2022. With 5,000 qubits, the Leap becomes Europe’s most powerful, commercially available quantum computer. Moreover, the D-Wave machine will be run in a hybrid configuration with a classical supercomputer.

McKinsey sees four major use cases for quantum computing in the near future. These are almost certain to shift as quantum computing itself grows. For example, the idea of a quantum internet was mostly conjecture as of January, 2022. Now, with quantum machines stable at up to five seconds, an internet working at quantum levels is something that can be not just posited, but worked on.

McKinsey’s use cases fall under:

  • Quantum simulation;
  • Quantum optimization and search;
  • Quantum factorization;
  • Quantum linear algebra for AI/ML.

Let’s look at each of these in turn. Since AI/ML is one of our favorite things, we’ll look at it last and look a little deeper into it.

Investing Into Quantum Computing

The investment market for quantum computing is growing rapidly. Quantum Insider determined that USD 3.2 billion of private capital investment had been announced for 2021 for the industry, with a third of that flowing in during 4Q21 alone.

Investments in 2021 include the use of a Special Purpose Acquisition Company, or SPAC, to bring Rigetti Computing to equity markets. SPACs are shell companies created for the purpose of acquiring an existing company and bringing it to the market. Because the shell company does not have any business of its own to report on, the regulatory burden on it is light and simple.
As a vehicle, SPACs are a common path for high-tech companies to gain access to the capital markets without the strictures and reporting requirements of a straightforward listing.  Also, SPACs have been a favored route to listing in particular for tech companies with hardware to show off – as in the space launch and electric vehicle industries. Bringing Rigetti out in this manner points to a certain level of maturity in the industry, and more are likely to follow.

Quantum Computing Use Cases

McKinsey sees four major use cases for quantum computing in the near future. These are almost certain to shift as quantum computing itself grows. For example, the idea of a quantum internet was mostly conjecture as of January, 2022. Now, with quantum machines stable at up to five seconds, an internet working at quantum levels is something that can be not just posited, but worked on.

McKinsey’s use cases fall under:

  • Quantum simulation;
  • Quantum optimization and search;
  • Quantum factorization;
  • Quantum linear algebra for AI/ML.

How these cases are realized in different industries. And since AI/ML is one of our favorite things, we’ll look at it last and look a little deeper into it.

Use Cases In Chemistry

Chemistry is seen as one of the first industrial verticals that will be transformed by quantum computing. Qualitative changes are expected across the full cycle of industrial chemistry, from faster design, to more precise formulation to more efficient production to streamlined logistics. Some of these aspects, especially later in the production chain, are shared with other industries, and among the quantum tools chemists utilize will certainly be AI/ML-driven. However, the nature of chemistry at the molecular level and smaller, makes quantum computing’s ability to solve multivariate problems a solution to many issues that were not solvable by classical computing given today’s hardware.

Chemical design will be quickened through quantum simulation, optimization, and AI. For example, it will become faster to design molecules  or polymers that meet even more specific requirements than are seen today. Classical computing can reach rough sketches in a limited number of cases, mostly involving lighter molecules and gasses. For weightier issues involving solids, heavy atoms or large molecules, classical computing is not up to the task.

The greater power to understand what exactly is being created when chemicals are designed will go beyond the efficiency or power of the chemical for its intended purpose. Many fields within chemistry will also benefit from the more accurate and faster ability to forecast how a given chemical will interact with the environment. For agriculture, like in medicine, this will be vital for affecting the consumer of the chemical (such as harmful insects) as well as those that are directly part of the food chain (such as birds that eat those insects) or come in contact indirectly (such as fish in bodies of water nearby).

The development of proteins is one area where chemistry is expected to benefit greatly from the use of quantum computers. Currently, classical computers fare poorly in their ability to handle protein structures. The greater accuracy and speed of quantum computers will improve the ability to predict interactions between proteins and other chemicals. This will affect our understanding of everything from how agrochemicals act in nature to the efficiency of detergents removing food stains on shirts.

Mixing these chemicals will become easier and safer with quantum simulation, thus making safety and regulatory concerns easier to meet. Moreover, environmental concerns could be lessened with better optimized production as well as more insight into the agents possible for breaking down hazardous waste in the first place.

Sooner Rather Than Later

The chemical industry stands to benefit from early-stage quantum computing because the level of processing power needed to model chemicals, though astounding in classical computational terms, is rather light for quantum computing.

Current practice in molecular chemistry, according to McKinsey, includes the use of functionals based on density functional theory (DFT) in a wide range of cases such as spectroscopy, the behavior of oxides, and the reaction of nanostructures to environmental pollutants. While the use of DFT is not without problems, it’s also a relatively small step for researchers in fields using such quantum mechanical modeling methods to step to quantum computing. This is in contrast to, say, the automotive industry, which at first glance is at more of a remove. Also, because DFT-derived figures need confirming through other means, the industry is accustomed to looking for outside verification. The promise of more accurate calculations from quantum computing can be checked via existing processes that are already commonplace.

One other problem that might be easier for chemists to solve than for their peers in other industries is finding someone with expertise in both quantum computing and the subfield of chemistry. Improving upon the limitations of DFT has led the chemical industry to become early adopters of quantum computing, and since researchers are already working at the quantum level, the learning curve is lower.

Just how much computing power does the chemical industry need for quantum computing to take hold? According to McKinsey’s 2018 report, machines running between 1,000 and 10,000 qubits should do the job, and that such machines should be available in the early-to-mid 2020s. In practice, The assessment was spot-on, as the first commercially available quantum computer in Europe running over 5,000 qubits was a D-Wave quantum annealer commissioned at Jülich UNified Infrastructure for Quantum computing (JUNIQ) in northern Germany in January 2022. JUNIQ’s director, Prof. Thomas Lippert, states that as far as he knows, the D-Wave at Jülich is the first to be tied directly to a supercomputer.

Use Cases In Pharmaceuticals

Quantum computing offers a way to cut the expense and development time required to bring a new pharmaceutical from inception to market. While the introduction of vaccines such as Moderna’s coronavirus vaccine may be the future, bringing a new pharmaceutical to market on average costs $2 billion and takes about 10 years. Testing across the development chain can be diminished as more certain techniques informed by quantum analysis give more certainty to the process.

Current pharmaceutical design, like chemistry in general, uses DFT based design and as such is starting from rough drafts of ideas. Initial hypotheses are more likely to be discovered via machine learning informed by quantum computing. Design and fine tuning comes from Computer Aided Drug Design (CADD) software informed by drug testing.

Quantum computing is expected to transform CADD in several ways. As with chemistry in general, the behavior of proteins should become more clearly defined, especially in terms of interactions with candidate drugs. Along with the greater accuracy, the ability to rummage through sets of libraries in parallel will speed up the search for compatible candidates while also enabling a wider search in the first place.

McKinsey sees search as only the beginning for quantum computing-fed CADD. As quantum computing matures, the discipline should be able to teach ML to hunt for new relationships between pharmaceuticals and target molecules. This hunt could include a wider range of more complicated molecules as well as being more automated, which should improve accuracy.

One stark example of the transformation possible in the pharmaceutical industry can be seen in target identification and validation. Quantum computing should be able to accurately model protein structures in three dimensions. In many cases 3D modeling is still beyond the reach of classical computers, but because quantum computing can take into account the actions of electrons, a breakthrough is possible. Complex interactions such as those between proteins are also likely to be describable for the first time.

Existing computation should improve as well. Google revolutionized the study of protein folding with its AlphaFold, which is on the cutting edge of modeling based on classical computing. The current version, AlphaFold 2, required weeks and up to 200 GPUs to train on 170,000 different protein structures, and subsequently days to come up with a solution for a structure.

Breaking away from classical computing models should not just speed up candidate discovery and confirmation but also expand the range of possible targets to include larger and more complex examples. Moreover, quantum computing, through ML, could be able to fill in missing data in rare diseases by providing ‘best guesses’ based on heterogeneous data, which would allow researchers to move forward and later tweak the results.

The promise of quantum computing in the pharmaceutical industry is likely to bear fruit later than for chemistry in general. This lies partly in the level of precision needed and the complexity of the tasks ahead, as well as in regulatory and health-and-safety issues. However, McKinsey sees the current state of quantum computing as one in which enterprises can begin to work with the technology. Once error correction is resolved within the field (probably, by 2030), however, a step change in the adoption of quantum computing in pharmaceuticals will be seen.

Use Cases In Financial Services

The financial services sector is likely to see sweeping changes as quantum computing takes hold. The changes are manifold and are not likely to happen all at once. Institutional investing is likely to be an early adopter, even for hybrid systems if the combination of speed and data informed by increasingly precise algorithms can result in a competitive advantage.

Risk managers are likely to follow institutional investors to the quantum level, as are those working in complex fields. By the time quantum computing reaches retail banking (if ever), other aspects of banking that do not attract so much attention can be affected. This includes forms of trading that rely not only on the speed of analysis, but also the connection between the investment bank and the trading platform. Recent advances in quantum stability point to possibilities for high speed communications.

Macroeconomics remains an open question. WIll the banking and finance industry’s macro analysts utilize the greater capacity of a hybrid system, or wait until quantum computing matures? The COVID pandemic brought surprises to all aspects of economics, from the initial shock to the K-shaped recovery to the Great Recession and extended bout of inflation. Better macro predictions would be of great interest, but this might not be a project for early adoption.

One aspect of financial services that comes as a second thought to most people is one that runs through many verticals, that is, encryption. Data safety is approaching an inflection point at which quantum computing lays waste to current security measures and new paradigms for keeping data safe are yet to be rolled out.

Use Cases In Artificial Intelligence

All of the verticals above make extensive use of Artificial Intelligence and Machine Learning, and quantum computing holds great promise in this direction. There are several advantages, which makes it worthwhile to look at AI, ML and neural networks separately.

As technological breakthroughs are overcome, the usefulness of quantum computing for contributing to the field of AI will spread. Quantum computing holds promise in furthering Artificial General Intelligence.  Quantum computing is already useful for speeding up the pace of machine learning models and for optimizing algorithms. Even before quantum computing matures beyond today’s decoherence and fault tolerance issues, it is providing answers in seconds to problems that would take classical computers years, and sometimes millenia, to solve. Utilizing quantum computing to Machine Learning in particular is already bearing fruit, and for research institutes, a quantum coprocessor (see sidebar) is now a reality.

Quantum AI In Practice

Google introduced TensorFlow Quantum(TFQ) in March 2020 as an open-source library for quantum machine learning. The project has partners, namely the University of Waterloo, X, and Volkswagen. TFQ integrates quantum and classical tools to enable users to create a unified workflow across the two types of computing:

  1. Convert quantum data to quantum: Quantum data can be represented as a multi-dimensional array of numbers called quantum tensors. TFQ works on these tensors to create a dataset for further use.
  2. Select the quantum neural network model to be used:  Quantum neural network models are selected depending on the quantum data structure.
  3. Sampling and Averaging: Measuring quantum states creates classical information. This information consists of samples from the classical distribution, though the values come from the quantum state. TFQ includes methods that average multiple runs based on the first two steps.
  4. Use classical methods to gather insight– With the data now in classical form, deep learning comes into play to further work with the data.

Low Hanging Fruit For Quantum Computing In AI

One of the ways that quantum computing can already further AI in general is to use it for creating better algorithms to use in classical computers doing AI/ML.

  • Learning: Classical learning models still need to be adapted to the quantum world. Doing so could bring improvements to the deep learning training process in terms of speed and accuracy. Furthermore, quantum computing can shorten the time needed to come up with the optimal solution set of artificial neural network weights.
  • Decision problem solving: Decision trees form the basis of classical decision problem solving, but there are limits. Problems can be too complex for a division by two, and when this happens, efficiency decreases. Quantum algorithms utilizing Hamiltonian time evolution can get around this issue.
  • Quantum search: While classical computing is faster than human-based search, there might be room for improvement. Quantum search should be able to beat classical computing when it comes to search, and the ramifications of this jump in speed will be felt in encryption, in particular.
  • Quantum game theory: Classical game theory is heavily utilized in AI. Extending game theory should bring about the realization of quantum communication and quantum artificial intelligence.

What Are The Critical Milestones For Quantum AI?

Quantum AI has come a long way, as in it actually exists. However, it still needs to reach some critical milestones to be reached before it can be called a mature technology. These milestones include:

  • Greater error correction and greater power quantum computing systems
  • Wider adoption of open-source frameworks
  • Developers! Developers! Developers! Building out an ecosystem
  • Quantum AI applications that drive use because of overwhelming benefits

Quantum Computing And Encryption

Quantum computing remained on the margins of mass media until the threat to security was raised. Suddenly, alarmist articles about how quantum computing was going to put our very passwords and with them, our personal data and banking information, into the cold hands of cryo-cyber-criminals.

Because of quantum computings capacity to work on problems in parallel, encryption that takes a classical computer years to break through can now be broken in a matter of minutes. The implications for security are tremendous, and a whole industry will need to overhaul itself.

Quantum-enabled encryption is not generating headlines, but it is under development. In a February 2022 article, CNET offered a vision of what the next generation of consumer encryption would look like. IBM and Thales, the authors point out, are calling it “post-quantum cryptography”.

According to this vision of the future, the scale of change to encryption would be on a par with fixing Y2K bugs or moving from IPv4 to IIPv6 Internet communications. It’s possible, but piecemeal and would take a multilevel approach. Furthermore, current-generation hardware is unlikely to suffice. While consumers would eventually need to replace their devices, governments, encryption industry players and businesses in general would need to come to terms with the new paradigm much sooner and would essentially drive adoption.

The U.S. National institute of Standards and Technology is already leading work to find resilient algorithms. A competition announced in 2016 for post-quantum algorithms is down to 15 candidate systems from an original 84 entries, according to a February 2022 article in Nature. The contenders include 9 public-key systems and 6 utilizing digital signatures.

In the meantime, an effort from the Chinese Association for Cryptologic Research posted the results of its own competition in 2020. The official Chinese post-quantum adoption comes under the purview of the Office of State Commercial Cryptography Administration. The relationship between the known official process and the group behind the 2020 announcement is still unclear.

Post-Quantum Cryptography For The Quantum Internet

The eventual NIST and Chinese-backed standards and systems are likely to compete in the industry. However, deciding on which way to go forward is only the beginning of the story.  While personal, corporate and government users will have to make decisions on how to adapt to the new technology, there remains a set of problems around the infrastructure required to make such solutions work in the first place.

Internet routing will need to be tested thoroughly to ensure that bottlenecks are not created or blockages occur given the new protocols involved. Worse yet, implementation even at the national level can be varied, and quantum-secure protocols could set off alarms in countries that have expectations of actively monitoring communications. ‘Protocol ossification’ is an issue.

What Is Quantum Annealing?

While quantum computing via quantum logic gates and registers from IBM and Google catch a lot of headlines, the quantum annealing machines produced by DWave are also pushing the envelope on quantum computing, albeit in a different direction.

Quantum Annealing Pros And Cons

Quantum annealing is a method of creating quantum computing that works with currently available technology at scales far beyond that of quantum logic gate-based technology.. However, like quantum logic gate-based machines, quantum annealers still need to work at temperatures close to absolute zero. Unlike quantum logic-based machines, though, annealers work with energy levels and specifically, energy minimums. Temperature control becomes a problem as more qubits are added. However, the problems are not insurmountable. For example, an article in physics.aps.org in 2017 pointed to the heat issue as a limiting factor with 1,000 qubit annealers. However, as previously mentioned, DWave successfully delivered 5,000 qubit annealers in 2021. And while quantum annealers are the least flexible of quantum computers, they are the easiest to build with qubit stability and quantity in mind.

In particular, quantum annealing is useful for solving optimization problems more effectively than traditional computers. Since optimization is a typical type of industrial issue, quantum annealing might become the most promising quantum technology for companies that are limited by classical computing in solving their urgent optimization problems.

How Does Quantum Annealing Work?

Quantum annealing is less noise sensitive than quantum gate model computing. As a result, it is easier to increase the number of qubits in use and with that, add more parameters when solving problems.

In quantum annealers, energy levels stand for each state, and by using superposition and entanglement, the lowest energy result can be found. This lowest energy state stands for the most likely or most efficient solution.

Why Does It Matter Now?

Quantum annealing already outpaces classical methods, which will help solve optimization problems in a variety of industries. Quantum annealing also breaks ground in terms of handling large quantities of qubits and issues such as heat dissipation.

With the amount of data produced on a daily basis increasing constantly, quantum annealing offers a solution not just to the volume of data but also the increasing diversity of that data. Science, mathematics and engineering stand to gain from it in the next few years. And while quantum annealing fits a narrower range of applications than other forms of quantum computing, the short term gains may help in the creation of more optimal quantum gate model computers.

Alternatives To Quantum Annealing

While quantum annealing has been developing rapidly in the last few years, there are still attempts to do the same processes by other means. Classical computing is still a viable option in some cases. This holds true particularly for simpler problems in which it might be easier to use existing classical computing.

Digital annealing is a way of simulating quantum annealing on a classical computer. Developed by Fujitsu, digital annealing uses an array of CMOS circuits to get around the environmental requirements of quantum computers. The result, according to Fujitsu, is not as fast as quantum annealing, but it is a more stable platform.

What Is Cloud-Based Quantum Computing?

Very few users of quantum computing would actually need to be physically located at the quantum computer itself to make use of it. Advances in classical computer networking have made remote access the norm rather than the exception. Likewise, the cost and intricacy of deploying a quantum computer and the systems required to support it make mass physical adoption unlikely at the current time.

How Does Cloud-Based Quantum Computing Work?

The need to connect would-be users to quantum machines has spawned an industry segment.  Companies such as Rigetti enable a workflow that goes like this:

  1. Developers work with a quantum machine image (QMI) via classical devices. Rigetti has created tools such as a python library called pyQuil using the company’s Quil language for developing and running quantum software applications.
  2. Quantum virtual machines (QVMs) are called for running the code developed. QVMs are used to run code testing and generate the waveform that runs on quantum processors.
  3. The QMI exchanges waveforms with the quantum processor to configure qubits via the waveforms.
  4. The QPU send the data to the QMI for processing, which then sends the results on to the classical computer.

Because cloud based quantum computing is already accepted as a norm for qorking on quantum computers, adoption of the technology should be faster than if individual machines needed to be purchased for each end user. Also, connecting quantum computing programming languages to existing means such as python libraries helps to build a wider set of bridges, and increase acceptance by users.

Top Cloud Quantum Computing Vendors

Companies from new-born startups to the oldest names in tech are developing quantum computing. While some technology giants are developing their own quantum systems all alone, most are engaged in strategic partnerships.

Here is some of the cloud-based quantum computing providers:

Microsoft Azure Quantum

Microsoft’s quantum compute offering includes tools as QDK and quantum script languages like Q# for quantum computing development. Microsoft’s partnerships include 1Qbit, Honeywell, IONQ, QCI on the quantum computing systems side. Azure Cloud is used to gain access to it’s partners’ quantum computers.

Microsoft also has its own quantum system, Station Q. The system uses what is called the topological qubit method for stable quantum bits. The main goal is to further the mass production of quantum computers.

IBM Q Experience

IBM’s quantum network was initiated in 2016, and is called IBM Q network. IBM has since become a leader in quantum computing. IBM Q cloud access is done through IBM’s open-source quantum software development kit, Qiskit.

Amazon Bracket

Amazon announced its entry into quantum computing in 2019 with Bracket. Leveraging its famous AWS platform, Amazon provides the entire system as QCaaS. Amazon also has physical infrastructure called Amazon Quantum Solutions Lab.

Google Quantum Playground

Quantum Playground includes a simulator with a user interface, a scripting language and 3-D quantum state visualization. Google is also involved in processor development and in 2019 announced that it had achieved quantum supremacy with its 54-qubit Sycamore processor. The claim remains a matter of contention.

Rigetti Forest

A quantum computing startup company from Day One, Rigetti has already raised a total of $190M in funding. Rigetti’s Forest product includes a tool suite for quantum computing with a programming language, development tools and example algorithms.

D-Wave Leap

D-wave produced the world’s first commercially available quantum computer. D-wave is a startup company, and it has raised more than $200M. D-wave’s QCaaS offering was accessible for free by means of its Leap cloud service during the COVID-19 pandemic.


Xanadu is focusing on photonic quantum cloud platforms. Current cloud-based offerings can provide access to up to 24 qubits. The company is working with Creative Destruction Labs, Scotia Bank, BMO and Oak Ridge National Laboratory (ORNL).

“QCaaS provides companies with the opportunity to grow accustomed to quantum methods even before all of the bugs are worked out. We get to grow with the industry from relatively early years. And with our finger on the pulse of AI/ML, we’re excited by how we see the intersect between quantum computing and artificial intelligence growing stronger with each development in the field. It’s a game changer for us, and at Proxet, we’re keeping a close eye out for how we can utilize it and how soon.”

Vlad Medvedovsky, CEO at Proxet (ex – Rails Reactor) – a custom software development company.

Quantum Computing As Of Early 2022

The quantum computing industry is growing rapidly. And like the sound barrier, seemingly impassable points become milestones within months or years.With that in mind, let’s take a look at announcements made commercially and in academia in the beginning of 2022.

Error Handling

Nature.com released three separate articles on January 19 alone examining methods that resulted in error-free operation at over 99%. The papers focus largely on improvements in spin control, and the increase in accuracy gained thereby. Reaching the 99% accuracy level for a single qubit has been achieved, but doing so in multiple-qubit environments had not been possible until now. The authors of the papers in Nature managed to reach the threshold required for classical computing error correction techniques to be utilized in a two-qubit gate system.

All three are silicon-based, which should enable uptake using current materials. Moreover, these error-free rates were possible in multi-qubit systems, which will help quicken commercial adoption. The gains themselves were made by looking at the accuracy of individual stages of the processing chain, and applying corrections at those points in particular. Though the three teams methods the teams utilized differed greatly. This diversification of routes should help foster adoption.

25-Qubit QPUs Available Commercially

Quantum computing hardware has largely been available in systems, with no possibility to build your own machine. This stands in stark contrast to the classical computing world, which has waves of computer building frenzies and a burgeoning retail market for CPUs. Dutch firm Quantware announced in March, 2022 that it had created a 25-qubit QPU that was now available. Moreover, orders for a processor would take only 30 days to fill. The industry is far from seeing websites such as QPUcheck.com for performance comparisons, but the day might be coming.

Considering that IBM has released its 127 qubit Eagle processor and is already looking toward 400 qubits and beyond, a 25 qubit processor might not seem like much. However, for companies in industries that can already make use of quantum computing, such as in chemistry, the ability to roll their own quantum computer and work in-house will start making the technology look like a tool that is in reach.

Preventing Decoherence

Quantum computing is expanding the boundaries of science and engineering in other fields as well. The U.S. Department of Energy’s Fermi National Accelerator Laboratory posted in February 2022 about the lab’s work in detecting impurities in a qubit. Materials science has been vital to the development of classical computing, especially in the development of the CPU. Fermilab’s scientists have shown that material science holds promise for quantum computing as well.

Fermilab worked with Rigetti to examine to see what kind of materials properties were affecting performance, and in what ways. The team found that material imperfections and impurities induced noise could affect the coherence, or ability to hold a quantum state, in a qubit. Removing those defects should help a qubit to cohere for much longer. Impurities such as oxygen, hydrogen, and carbon, among others, were found at the atomic level.

The type of work done to reach an understanding of the impurities was out of the ordinary, even for Fermi. “The tools needed to perform this characterization are not only specialized and expensive, but also require experienced scientists to acquire and analyze the data,” noted Anthony McFadden, a qubit fabrication expert at National Institute of Science and Technology.

The main tool, a time-of-flight secondary ion mass spectrometer, was repurposed from analyzing atomic level imperfections in the accelerator cavities used in particle accelerators.

Market Forecast To 2026

Research and Markets released a report in February 2022 stating that the global quantum computing market would reach $1.6 billion in 2026, which is a 33.2$ compound annual growth rate and up from $390.7 million in 2021.

In particular, Research and Markets saw the superconducting qubits segment (which includes IBM and Google) growing from $144.9 million in 2021 t0 $612.9 million in 2026. The trapped ions segment (which includes D-Wave) should see growth from $131.1 million in 2021 to $550.7 million in 2026.

Industry trends such as partnerships and joint ventures as well as QCaaS are likely to continue, according to the report. However, the need for more quantum computing machines themselves is likely to rise over the next four years.


Quantum computing is a quickly growing form of computing that offers the prospect of quickly solving currently untouchable problems, while also creating disruptions to communications and security. The data-intense response to the COVID pandemic and current state of quantum computing hardware create an openness to the use cases for utilizing the optimization and other problem solving capacities of current generation hardware as well as the need to invest in passing developmental milestones. Classical computing will continue to play a dominant role in the IT aspects of consumers’ and industries’ daily routines, but quantum computing, like supercomputers, will have their place as well.

Related Posts