2020/02/03 Yole -Quantum technologies are speeding up to commercialization, but the road to a quantum computer is still tortuous

The post-Snowden world is ahead!

Data and its conduits are today an unprecedented arena of political struggle, centered on surveillance and privacy. At the same time, Big Data will require new ways to search and process. In this context, it is likely that new ways of computing will be required, but also sensing and ultra-secure transmission. Also, the world today faces huge environmental, security and healthcare challenges that need new solutions. The increasing amount of data and ever-growing number of parameters to be considered will challenge the classical semiconductor industry.

Moore’s Law, which has governed the microprocessor industry for over 50 years, could soon reach its limit. Quantum computers, based on the principles of quantum mechanics, are a real disruption and are expected to overcome the physical boundaries of present-day computing. They will open up new opportunities, particularly in intensive computing. However, quantum computing is not a continuation of Moore’s law or Artificial Intelligence. It is a totally new and disruptive approach based on different hardware and software.

Dead or alive or both? What is quantum?

Strange things occur in the quantum realm. This realm is at the atom scale, and involves light-speed phenomena. This is where we find the basic building block of a quantum computer, the qubit. A qubit operates in a multidimensional universe, with its eigenstates corresponding to the surface of a so-called Bloch sphere, while its logical states correspond to the poles of this sphere. More simply said, a qubit is not binary. It does not simply encode a 1 or a 0 as a bit does, but is a superposition of 1 and 0, much like Schrodinger’s cat is both dead and alive. Thus, a set of n qubits will encode a superposition of 2n possible quantum states. In general, a quantum computer with n qubits can be in any superposition of up to 2n different states. This compares to a normal computer that can only be in one of these 2n states at one time and opens up much more powerful computation possibilities. But we should remember what Bill Philips, Nobel Prize physicist said: “A quantum computer is as different from a classical computer as a classical computer is from an abacus.”

A new quantum revolution ahead

Quantum technology is important as it addresses secure communications and database management linked to national security. But it also addresses key industrial simulation and optimization challenges in chemicals and materials research, logistics, financial services, healthcare and life science, manufacturing, drug discovery, protein structure prediction, investment risk analysis, feedstock management, vehicle routing and network optimization. Market potential is accordingly huge. Yole Développement (Yole) forecasts that the total market value for quantum technologies, including computing, cryptography and sensing, will grow from about $480M in 2018 to $3.2B in 2030, with a 17% Compound Annual Growth Rate (CAGR). This quantum computing hardware market value is worth about $30M today, mostly being quantum annealers, and will grow to $650M in 2030. Yole estimates that Quantum as a Service (QaaS) will be worth $1.37B in 2030!

Powerful countries are aware of the possibilities that quantum technologies will open up. Quantum technology will be a national priority for the different countries and top world powers have invested in quantum projects. China has announced a $10B investment, the US a $1.2B investment over five years, and Europe announced €1.2B over ten years. France just announced a very ambitious national plan for quantum technologies over five years.

Not just quantum computers

Quantum technologies include computing, of course, but also cryptography, sensing and software. Quantum computing is today attracting most interest from the R&D and investor community but other quantum fields are benefiting from it. Quantum communication is a more mature technology, at least for short distances today, while quantum sensors and clocks are still small markets. Both cryptography and sensors are nevertheless benefiting from the current excitement around quantum computers. We see more and more future uses for quantum sensors for geophysics, for example.

Quantum computers come in two types. Quantum annealers like D-Wave’s have been commercialized as quantum computers for optimization problems for a few years now. Meanwhile, “universal” quantum computers can solve any kind of problem. Today the market is small as few companies can afford the huge development cost associated with the development of a quantum computer. D-Wave has already delivered four generations of quantum annealers over seven years, with the latest generation handling 2,000 qubits. Other companies developing universal quantum computers include: Google, IBM, Rigetti, IonQ, Intel and ATOS (see figure). In September 2019, Google’s claim to have demonstrated quantum supremacy with 53 qubits, corresponding to 253 states, created a huge buzz. And competition is becoming tougher! In December 2019, Intel announced its new cryogenic control chip codenamed Horse Ridge that will speed up development of full-stack quantum computing systems. Intel’s Horse Ridge chip was co-developed by Intel Labs and QuTech, a joint venture between TU Delft and TNO and is made using Intel’s 22nm FinFET process technology. This is a great achievement as it can interconnect and control multiple qubits at the same time, an essential capability required to build a large-scale commercial quantum system.

What next?

Quantum technologies are at the crossroads of numerous applications and fields, including research, material science, photonics, semiconductors, engineering, software and education. The technologies are still at an early stage but with huge potential. However they raise numerous questions: Which qubit technology will be favored? Which business model has to be defined? Will we be able to use semiconductor processes for manufacturing quantum chips? Will we be able to reduce cost?

The road to the realization of a quantum computer is long. Qubit control in terms of fidelity, coherence and scalability is a challenge. Cryogenic control technology is also critical for the access to this technology. And we still lack a common quantum language for all the different quantum computers currently developed. Yole believes the real value of quantum computing is in the full range of services it will offer: developing new materials, new drugs, optimizing smart grids for energy and telecommunication networks, to simulate traffic – this is where Yole forecasts the highest market value.