This resource covers the future of computing in the post-transistor age, and the technical hurdles inherent in the pursuit of quantum computing.
Quantum computing — considered to be the next generation of high-performance computing — is a rapidly-changing field that receives equal parts attention in academia and in enterprise research labs. Presently, Google, IBM, and Intel are independently developing their own implementations of quantum computers, as are startups such as D-Wave Systems. TechRepublic's cheat sheet for quantum computing is positioned both as an easily digestable introduction to a new paradigm of computing, as well as a "living" guide that will be updated periodically to keep IT leaders informed on advances in the science and commercialization of quantum computing.
SEE: Ebook—IT leader's guide to the future of quantum computing (Tech Pro Research)
- What is quantum computing? Quantum computing is a developing technology, which scientists anticipate will provide faster computational solutions to problems currently handled by supercomputers.
- Why does quantum computing matter? Theoretically, quantum computers could be used to crack RSA cryptography, which is commonly used across the internet.
- Who does quantum computing affect? Presently, primarily researchers working in quantum physics, though advances in quantum computing are anticipated to influence other "fuzzy logic" disciplines, such as artificial intelligence and machine learning.
- When will quantum computers be released? Systems limited to a specific type of quantum computation called digital annealing are commercially available, though there is not yet a clear benefit compared to traditional computers.
- How do I get a quantum computer? Multiple vendors offer cloud-based access to quantum computers. Even though purchasing a system outright is possible, it is likely cost prohibitive, as current systems are only useful for specialized workloads.
What is quantum computing?
Quantum computing is an emerging technology that attempts to overcome limitations inherent to traditional, transistor-based computers. Transistor-based computers rely on the encoding of data in binary bits—either 0 or 1. Quantum computers utilize qubits, which have different operational properties. While it is possible to encode binary data in a qubit, the natural state of a qubit is essentially superposition. This property allows qubits to have values of 0 and 1 (or values between 0 and 1) simultaneously. Likewise, because of the properties of quantum physics, multiple measurements of qubits in identical states will not return identical results. Qubits can also contain up to two bits of binary data as part of a process called superdense coding.
Using quantum computation, mathematically complex tasks that are at present typically handled by supercomputers — protein folding, for example — can theoretically be performed by quantum computers at a lower energy cost than transistor-based supercomputers. While current quantum machines are essentially proof-of-concept devices, the algorithms which would be used on production-ready machines are being tested presently, to ensure that the results are predictable and reproducible. At the current stage of development, a given problem can be solved by both quantum and traditional (binary) computers. As manufacturing processes used to build quantum computers is refined, it is anticipated that they will become faster at computational tasks than traditional, binary computers.
Further, quantum supremacy is the threshold at which quantum computers are theorized to be capable of solving problems, which traditional computers would not (practically) be able to solve. Practically speaking, quantum supremacy would provide a superpolynomial speed increase over the best known (or possible) algorithm designed for traditional computers. Theoretically, this can be demonstrated using Shor's algorithm for prime factorization, which would provide such a speed increase when performed on a quantum computer, as factoring is thought to be generally hard with traditional computers (though, this is not proven, in the scientific sense of "proof").
A research paper published in Science in October 2018 titled "Quantum advantage with shallow circuits" tested a variant of the Bernstein-Vazirani problem, in which researchers proved that a quantum computer with a fixed circuit depth will outperform a classical computer used to compute the same problem. While this does not itself establish quantum supremacy, it does demonstrate the potential of quantum computers as refined designs increase the number of qubits, and the length of quantum coherence, allowing for more complex calculations to be performed.
- The 5 basics of quantum computers (TechRepublic)
- Video: How quantum computing will better simulate the real world (TechRepublic)
- Photos: The world's 25 fastest supercomputers (TechRepublic)
Why does quantum computing matter?
Theoretically, advancements in quantum computing would lead to a breakthrough in integer factorization. If integer factorization became trivial to perform, the integrity of commonly used encryption systems would be shattered, allowing any individual, organization, or government with access to quantum computers the ability to brute-force decryption keys, with which locked devices or encrypted archives can be made accessible. Because of concerns in the cybersecurity community about the viability of quantum computers in breaking encryption, research into lattice-based cryptography — which is thought to not be susceptible to being broken by quantum computers — has increased.
To that end, on January 2014, reports indicated that the NSA has spent $79.7 million on a program titled "Penetrating Hard Targets." As part of this program, research was conducted to build "a cryptologically useful quantum computer." The documents cited in this report indicate that the NSA has not been appreciably more successful than other researchers. Likewise, the National Institute of Standards and Technology (NIST) published a request in December 2016 asking for public input on how to protect computers from the threat of quantum computers being used to crack encryption.
There is no consensus on when quantum computers will be capable of cracking encryption. In a May 2018 interview with TechRepublic, Bob Sutor, IBM Research's Vice President of Cognitive, Blockchain, and Quantum Solutions, estimated that quantum computers are 30-40 years away from breaking traditional cryptographic algorithms. The same month, IBM Research director Arvind Krishna warned that "Anyone that wants to make sure that their data is protected for longer than 10 years should move to alternate forms of encryption now."
Quantum computing is also anticipated to have other meaningful impacts outside of the field of cryptography. Because of the nature of quantum computation, they are uniquely well suited to so-called "optimization problems," where an exponential number of permutations to evaluate exist. In an interview with TechRepublic's Nick Heath, Andy Stanford Clark, IBM CTO for UK and Ireland provided an example: "If... you're optimizing the lengths of aircraft routes, or optimizing the layout of spare parts for a rail network, something where there's 2n possibilities and you've got to try each out in order to find the optimal solution. If you had a 2100 problem, which would be basically impossible to solve on a classical computer, with a 100-qubit quantum computer, you'd be able to solve it in one operation."
- How the industry expects to secure information in a quantum world (ZDNet)
- Quantum computers and the unbreakable lattice (ZDNet)
- Microsoft deepens University of Sydney quantum research partnership (ZDNet)
- How quantum computing could unpick encryption to reveal decades of online secrets (TechRepublic)
Who does quantum computing affect?
Research into quantum computing is driving a great deal of investment from universities, IT companies, and venture capital. Multiple public-private partnerships have sprung up as businesses work with research departments in universities to find use cases where quantum computing can be applied to existing business operations.
The IBM Q Network is the largest of these, with participating universities including North Carolina State University, Melbourne University, Oxford University, and Keio University, and participating companies including Samsung, JPMorgan Chase, Mitsubishi UFJ Financial Group, Mizuho Financial Group, and Mitsubishi Chemical.
Others include a collaboration between the Australian firm Silicon Quantum Computing, and France's national research and development (R&D) organization, the Commissariat à l'Energie Atomique et aux Energies Alternatives (CEA).
- CIOs watch out: Here are 7 disruptions you might not see coming (TechRepublic)
- China sends 'unbreakable code' from quantum satellite to Earth (TechRepublic)
- How quantum computing could unpick encryption to reveal decades of online secrets (TechRepublic)
- Google's quantum computer inches nearer after landmark performance breakthrough (ZDNet)
- Google looks to build quantum computer chips (ZDNet)
- IBM gets grant to help promote quantum computing research (ZDNet)
- The $50m Intel investment that's bringing quantum computing a little closer to the real world (ZDNet)
When will quantum computers be available?
There are two answers to this question: Now, and substantially far in the future. The Canadian company D-Wave Systems currently sells a quantum computer named the D-Wave 2000Q, however, there are significant caveats with that offering. D-Wave advertises this system as having 2000 qubits, though differences in D-Wave's definition of qubit relative to the rest of the quantum computing industry make this measurement not practically useful.
Further, the systems sold by D-Wave are designed specifically for quadratic unconstrained binary optimization, making them unsuitable for integer factorization required for cracking RSA encryption systems. Additionally, the D-Wave 2 (second-generation system) was found to not be faster than a traditional computer. For more information on how D-Wave products differ from other quantum computers, TechRepublic has a separate cheat sheet for D-Wave's quantum computer technology.
Likewise, Fujitsu offers a "quantum inspired" digital annealer, which is a traditional transistor-based computer designed for quantum annealing tasks, like D-Wave's quantum computer. However, Fujitsu does not market this system as a true quantum computer, as the traditional transistor-based design allows it to operate at room temperature without requiring helium-based cooling solutions, as well as making it resistant to noise and environmental conditions which impact performance in quantum computers.
In a general sense, it is possible that quantum computing may be a viable alternative in the future to current transistor-based solutions, though non-trivial encumbrances in fabrication and mass-manufacturing must be addressed for this to become a viable technology for mass industry adoption. Among these encumbrances are the difficulty of building computers which scale to multiple qubits, the ability to initialize qubits to a predictable value, and easing the means by which qubits can be read.
- Microsoft is hiring former Qualcomm engineers to work on its quantum computing team (ZDNet)
- Microsoft just upped its multi-million bet on quantum computing (ZDNet)
- Google expands work in quantum computing with 72-qubit Bristlecone processor (TechRepublic)
- Silicon Quantum Computing launched to commercialise UNSW quantum work (ZDNet)
- Google's quantum computer can blitz a normal PC - but it's not ready for production just yet (ZDNet)
- The latest in quantum computing: 10ft tall, 2,000 qubits, $15m price tag (ZDNet)
How do I get a quantum computer?
Quantum computing resources are widely available via cloud services, with vendor-specific frameworks. Presently, offerings are available from IBM Q (via Qiskit), while Google has introduced the Cirq framework, though it does not presently have a cloud offering in general availability. D-Wave Leap allows approved developers to conduct quantum experiments for free. Similarly, Fujitsu offers cloud access to their digital annealer system.
For buying systems outright, D-Wave's 2000Q system costs $15 million (Notable buyers include Volkswagen Group and Virginia Tech.). A quantum computer is not something you are likely to find at your local big-box store. However, if your workloads are more general, building and buying a POWER9 deployment is likely a better value at present. Oak Ridge National Laboratory's SUMMIT supercomputer is a POWER9 and NVIDIA Volta-driven system planned at 4600 nodes, with a computational performance in excess of 40 teraflops per node.