D-Wave Systems’ quantum computers have the potential to solve problems that the fastest supercomputers available today just can’t crack.
But claims of performance superiority have been criticized as premature by some academics, because D-Wave’s machines have yet to definitively prove themselves in the real world.
You’ll find everything you need to know about the D-Wave quantum computers in this “living” article, which will be updated over time.
- What are D-Wave quantum computers? They are machines that solve a specific class of problem by exploiting the counter-intuitive behavior of matter at the atomic level.
- Why do D-Wave quantum computers matter? Because of their potential to tackle problems that would be impossible for conventional computers to practically solve, with applications ranging from bioscience to cyber security.
- Who do D-Wave quantum computers affect? Mainly large organizations with deep pockets, as each D-Wave machine costs $15m. However, while early customers are restricted to the likes of Lockheed Martin, the US government, Google, and Nasa, D-Wave also provides access to its machines via a cloud service.
- When are D-Wave quantum computers happening? D-Wave plans to continue to develop its quantum processors, is backed by high-profile investors and continues to sell machines to the occasional early adopter.
- Who are D-Wave quantum computers’ competitors? Google’s Quantum AI Lab has developed a 72-qubit quantum processor called ‘Bristlecone’, while IBM has pledged it will build a 50-qubit quantum computer within “the next few years”.
What are D-Wave quantum computers?
D-Wave’s quantum computers use very different technology to that found in everyday computers.
The latest D-Wave machine is 10-feet tall, costs $15m and tackles problems using “quantum transistors”, tiny loops of niobium cooled to close to absolute zero (-459.6F) by liquid helium.
This exotic architecture is necessary for the D-Wave chips to exploit quantum phenomena, the counter-intuitive way that matter behaves at an atomic level.
D-Wave has gone to these lengths because it believes its processors have the potential to massively outstrip classical computers when it comes to solving a particular class of problem.
However, these processors are also fundamentally more limited in the breadth of problems they can tackle than the general-purpose computers in use today, with D-Wave’s systems only able to handle a specific type of computation.
Even though D-Wave calls its systems “computers”, John Morton, professor of nanoelectronics and nanophotonics at UCL, said that just as a calculator isn’t a computer, the D-Wave system isn’t a universal quantum computer.
“A calculator solves a very specific set of problems. Lots of people use it, and you can use calculators across many different industries,” he said.
“So when D-Wave shows you many different industries that might use a D-Wave machine, there may be many areas that it can be used in, but it remains a specialized device.”
Why do D-Wave quantum computers matter?
What makes D-Wave systems interesting is their potential.
Despite being at an early stage of development, organizations such as Nasa and Google, Lockheed Martin and Los Alamos National Laboratory have shelled out the $15m it costs for one of D-Wave’s machines.
That’s because D-Wave processors may eventually massively outperform classical computers when it comes to solving a particular class of mathematical problem called unconstrained binary optimization. A very simple example of this type of optimization problem might be the challenge of drawing up a plan for a house that comes as close to your dream spec as possible, while staying within your budget.
As an example of just how much D-Wave’s machines might one day outclass conventional computers, in 2015 a test by Google found that the D-Wave 2X processor was 100 million times faster than a classical processor running a similar operation.
The significance of that 100 million speedup was disputed, on the grounds of the tests being synthetic and massively favoring the D-Wave processor. However, D-Wave said the test was meaningful because it demonstrated that D-Wave’s fundamental approach was sound, that the chips were capable of exploiting the phenomena called quantum tunnelling to help perform calculations.
This tunnelling is necessary for D-Wave processors to carry out quantum annealing, the process of finding the minimum energy state for a system of particles, which is useful in modelling and solving the class of optimization problems mentioned above.
However, much of the promise of the D-Wave systems lie in their future, and there are some who cast doubt on whether that potential will ever be reached.
D-Wave’s most vociferous skeptic is probably Scott Aaronson, a computer science professor at the University of Texas in Austin.
While D-Wave has repeatedly claimed its tests show its processors’ superiority over classical counterparts for cracking certain problems, Aaronson says that, barring D-Wave’s most recent performance claims which are still being examined, past assertions of performance leads have been debunked, and that in each instance a different classical approach was found that “eliminated the claimed gap“.
D-Wave is yet to definitively demonstrate the so-called quantum supremacy of its processor, the ability to perform a calculation at a speed that classical supercomputers have no hope of matching.
However, Google’s director of engineering, Hartmut Neven, has said the technology giant is “optimistic” that the “significant runtime gains” demonstrated by using D-Wave in testing will “carry over to commercially relevant problems as they occur in tasks relevant to machine intelligence”.
Who do D-Wave quantum computers affect?
The binary optimization problem that D-Wave processors excel at has practical applications in a variety of areas.
D-Wave says that its processors have been used in the financial sector for trading trajectory optimization, to work out how proteins fold in bioscience, to create filters for lists that never miss a potential match — useful for security services checking terrorist watchlists, for spotting cyber security threats in online traffic, and for development of binary classifiers in AI and for computer vision.
Of particular interest to D-Wave is the potential for its processors to be used to carry out unsupervised machine learning, where unlabelled training data is fed into a neural network and the machine learns by identifying patterns.
D-Wave has already experimented with machine learning on the chip, setting up a Boltzmann machine, a type of stochastic recurrent neural network, as well as a “Quantum Boltzmann machine”, which Colin Williams, director of business development and strategic partnerships at D-Wave, said is ‘fundamentally different from previous machine learning models’ and could eventually allow a machine to ‘generate new data that is statistically indistinguishable from the kind of data on which it was trained’.
Williams predicts that eventually, a D-Wave-based machine learning model could be trained to produce new and convincing works of art in the style of the painter it was trained on or to replicate human-like speech.
D-Wave has launched a spin-off company called Quadrant to focus on how its quantum machines could be applied to deep learning, specializing in training machines using only small amounts of labelled data.
D-Wave doesn’t see its machines as a replacement for conventional computers, but as a complement, used to handle particular tasks before handing off work to a classical system.
When are D-Wave quantum computers happening?
Given the level of interest in D-Wave, it seems highly likely it will continue to release new machines for the foreseeable future.
D-Wave has raised millions of dollars in funding from various high-profile investors, including investment bank Goldman Sachs, In-Q-Tel (the investment arm of the US Central Intelligence Agency), Bezos Expeditions (the investment arm of Amazon founder Jeff Bezos), and BDC Capital, Harris & Harris Group, and DFJ.
Even though D-Wave has only sold a handful of its quantum computers, it continues to attract buyers for its machines and to generate interest in its technology. The carmaker Volkswagen has been working with D-Wave since 2017, using its systems to model traffic flows and develop more efficient batteries for electric cars.
In D-Wave’s view, the core technology at the heart of the chip has been demonstrated to work, and realizing its promise of quantum supremacy requires adding more qubits (quantum bits) to the processor and making these qubits more densely connected.
Towards the end of 2016, D-Wave launched its first 2000 qubit “quantum computer”, the 2000Q, which as well as doubling the number of qubits, introduced architectural improvements that the firm claimed helped speed up certain calculations 1000-fold over its predecessor, and 2,600x over classical computers.
US Department of Energy researchers at Oak Ridge National Laboratory in Tennessee will have cloud access to a D-Wave 2000Q system, allowing them to explore hybrid computing architectures, with a view to accelerating the development of software designed to run exascale systems.
D-Wave continues to find new customers for its systems. A D-Wave 2000Q system was also installed at the Quantum Artificial Intelligence Lab run by Google, NASA, and Universities Space Research Association early in 2018.
Beyond the 2000Q, D-Wave has a design for a “next-generation chip” with a “fundamentally new topology, based on all the lessons we’ve learnt”, which reportedly will both increase connectivity between qubits significantly and allow D-Wave to surpass the 10,000-qubit limit in its existing processors. D-Wave announced it had completed fabrication and testing of a working prototype next-generation processor early in February 2018.
Who are D-Wave quantum computers’ competitors?
As mentioned, D-Wave isn’t a universal quantum computer and UCL’s Morton predicts they may not exist until the 2030s.
However, many major tech companies are researching and developing universal quantum computer technology.
In March 2018, Google’s Quantum AI Lab showed off a new 72-qubit quantum processor called ‘Bristlecone’, which it says could soon achieve ‘quantum supremacy’ by outperforming a classical supercomputer on certain classes of problems.
IBM has also pledged it will build a 50-qubit quantum computer that will be commercially available within “the next few years”, in January 2018 the chipmaker Intel announced its own 49-qubit quantum chip and Microsoft is investing significant funding in quantum computing research.