Summary

• D-Wave has a significant first-mover advantage with commercially useful quantum annealing technology, unlike competitors who are decades away from viable quantum gate computers.

• A Recent capital raise at a higher share price reduced forecast dilution, and changes to its business model, including premium pricing, improve the margin outlook.

• D-Wave's sale of its first quantum computer and record bookings suggest it will exceed revenue forecasts, perhaps doubling medium-term revenue targets.

• Despite market volatility, D-Wave's unique position and growing commercial traction make it a strong buy, poised to outperform other quantum computing stocks.

gece33

In this article, I will argue that D-Wave (NYSE:QBTS) has developed a significant and sustainable competitive advantage over its immediate competitors IONQ, RGTI, QUBT, the major players in the quantum computing world like MSFT, GOOG, and IBM, as well as the many private companies developing quantum computing hardware.

I will present evidence that D-Waves' competitors are decades away from delivering a commercially useful device. In contrast, the D-Wave device is already commercially useful and used by many companies.

I believe D-Wave has a decade-long first mover advantage, that its products will make a significant difference to the cost structure of large businesses and that D-Wave is substantially undervalued.

Quantum computing is beset by two seemingly intractable problems: scaling and error correction. I do not believe D-Wave's competitors will be able to solve these problems in the foreseeable future, if at all, and will not produce a commercially useful quantum device before the end of the 2030s. This does not mean they will not make any computers or generate any revenue, just that the revenue will be from organizations researching quantum computers, not using quantum computers to solve real-world problems.

D-Wave quantum computers are not like the others. They use annealing technology that is already commercially useful and gaining traction with a growing number of customers. They have just sold their first quantum computer for perhaps $15 million, which suggests they will beat 2025 revenue forecasts by more than 100%.

The recent share price jump allowed them to raise capital at a far lower level of dilution than I had expected, significantly improving the balance sheet and taking away the main concern I raised in my previous D-Wave article.

This is my fourth article on D-Wave and my 8th on Quantum computing companies. It has been a very profitable sector for me, I lost 19% on my first investment in IONQ but made over 600% on D-Wave after my October article on the company.

Due Diligence:

I interviewed the CEO and CFO of D-Wave on Jan 10, 2025 to get an update on what had happened since my previous interview with them in October. The interview forms the basis of this article, which updates my previous one.

Types of Quantum Computer

There are two types of quantum computers. Quantum Gate Computers and Quantum Annealing Computers.

D-Wave is unique as it makes a quantum annealing computer, all other manufacturers are trying to develop quantum gate computers.

This is a short recap of the technology, which I have explained in detail in previous articles.

Qubits as the Building Blocks: Instead of classical bits (which are either 0 or 1), quantum computers use qubits. Qubits can exist in a superposition of both 0 and 1 states, offering a vast increase in computational power. The choice of 0 and 1 is arbitrary, it could be x and y or -1 and 1 or even North and East. Superposition allows quantum computers to explore many possibilities in parallel and allows for each qubit to store an infinite amount of information.

I have used the image below to help people understand superposition. The value a qubit holds can be thought of as a point on an Argand diagram. The orange dot on the circle would be the value of the qubit when measured, it is in a superposition, or combination, of the states x and y because the circle has an infinite number of points the qubit can return an infinite number of values when measured.

Argand Diagram (Author)

Entanglement: Qubits can become entangled, meaning their states are linked, regardless of the distance between them. This creates powerful correlations that can be exploited for computation.

Quantum Gates are Operations: Quantum gates are operations that manipulate the states of qubits. They are analogous to logic gates in classical computers (like AND, OR, NOT). Each gate moves the orange dot on the Argand diagram above to a different point on the circle.

Quantum Algorithms: A sequence of quantum gates applied to qubits. This defines the specific computation the quantum computer will perform.

Measurement: At the end of the computation, the qubits are measured. This destroys the superposition state but gives an answer to the algorithm being run.

What's the Problem?

In a nutshell, Qubits will not stay where you put them; after a gate operation, the qubit may or may not go where it is supposed to and may or may not stay where you put it. Qubits interact with other qubits, any heat, radiation, passing photons, or anything at all, and any interaction causes them to change value. They even change values for no reason we understand; when you measure them, they lose all of their value completely. They are inherently unreliable, and it has not been proven that this unreliability will ever be eradicated. A paper by Jack H Hiller, a regular Seeking Alpha commentator, suggests that the decoherence problem may never be solved and explains that we still do not yet understand the signaling process in entanglement despite the proof of its existence.

Companies developing quantum gate computers are attempting to solve the error problem by grouping together qubits into logical qubits and developing error-correcting software applications. Microsoft has been leading the way, but Google has made great strides recently. Still, I believe that these two leaders are decades away from a solution, and the smaller start-ups are even further behind.

Quantum Annealers

Quantum Annealers do not work in the way described above. Importantly, they do not apply quantum gates or attempt to move the qubit to another value. As a result, they are not hampered by the same error problems.

A D-Wave CPU uses superconducting qubits, small loops of superconducting wire with a circulating electrical current and a corresponding magnetic field. The current could flow in either direction (called the spin), and as it is a quantum particle, it could be in a superposition of the two directions. D-Wave applies magnetic fields to the qubit, and like all quantum particles, it will move to its lowest energy state by changing its superposition. That state can be measured and will return either spin up or spin down depending on the direction the current is traveling.

D-Wave entangles their qubits, linking them together using a coupler. The coupler can correlate the qubits with the same or opposing spin. The D-wave CPU controls the magnetic field applied to each qubit and the links provided by the couplers. Once the system is set up, the qubits will move to the lowest energy position in the connected system.

The D-Wave machine has limited application, it can solve two types of problems.

Optimization problems: for instance, the minimum number of miles a delivery driver has to travel to deliver all his parcels. In this case the magnetic fields, couplers and qubits would be set up to reflect the miles between each address and the lowest energy state of the qubit system would represent the lowest number of miles in the problem.

Machine learning problems: Machine learning, or AI as we now call it, looks for the best solution to a probabilistic model. You give the machine a lot of data to learn from and then ask it to fill in some gaps, when the sample is an energy-based distribution the D-Wave architecture is particularly good at finding the lowest energy solution.

Although these are only two use cases, they answer every company's two most important questions. How can you do this for less? And how can you do it better? The larger the business, the more important this is. Large manufacturing companies could develop a significant cost advantage over their competitors by using the D-Wave machine, and I covered in my previous article that many such companies are trying to do exactly that.

NISQ

The share price fall in January of all the quantum stocks may be the market realizing that the quantum computing companies will not generate meaningful revenue for an extended period because NISQ computers cannot solve useful problems.

We are in the era of Noisy Intermediate Scale Quantum computers. The term NISQ was coined by John Preskill in 2018. Preskill is an American theoretical physicist and a leading quantum information science and computation scientist. He is the Richard P. Feynman Professor of Theoretical Physics at Caltec and the Director of the Institute for Quantum Information and Matter.

IONQ and RGTI have both suggested that they are exploring the possibility of using NISQ computers to solve real problems today. Indeed, the CEO of IONQ recently predicted that revenue for his company would hit $1 billion of revenue on the back of these use cases.

After the IONQ CEO first made this statement in the Q3 earnings call, I spent a long time reviewing every published scientific paper I could find on the application of NISQ computers, and could not find a single one that suggested they could solve a useful problem. During our meeting, I relayed this to the D-Wave CEO, and he replied, "that is because there isn't any". He is not the only one who shares my view that NISQ computers cannot solve commercially useful problems, and that a quantum gate machine does not currently exist that can solve problems beyond those solvable by existing classical computers.

On Dec 14 2024 John Preskill posted his keynote address from the Q2B 2024 conference; in that address, he made two important remarks in the context of this article

NISQ already has noteworthy scientific value. But as of now there is no proposed application of NISQ computing with commercial value……..Furthermore, currently there are no persuasive theoretical arguments indicating that commercially viable applications will be found

Jensen Huang, CEO of Nvidia (NVDA), said at the recent CES 2025 show that he believes useful quantum computers are 15 to 30 years away. Huang focused on the problem of scaling quantum computers, saying they need to reach 1 million qubits before they become useful.

In his keynote address, John Preskill coined a new term FASQ, Fault-tolerant Application Scale Quantum computers, machines that can run commercially useful applications. He said these machines are "a rather distant goal". Preskill again maintained that useful machines will need to execute a million operations and need an error rate less than anything we expect to achieve anytime soon.

In my Microsoft article, I explained why current NISQ computers have a near-zero chance of getting the right answer after performing a million operations. (The actual probability was 0 point three hundred and thirty-five zeros 3)

Regular readers of my Quantum articles will recognize this one million figure, as I have repeated it in almost every article. It remains the mathematically proven number of qubits needed to run the vast body of quantum algorithms developed over the last 30 years.

In response to the intervention of Jensen Huang, which seemed to start a share price fall, the IONQ CEO reiterated that they are focused on finding applications for their current computers and as they scale up from AQ36 to AQ256 they will be able to tackle increasingly difficult problems.

We know that no useful application has been found for IONQs NISQ computers and there is no theoretical argument suggesting any can be found , AQ265 will still be a NISQ and I do not believe it will be able to solve any problem a current classical computer can.

Additions to the D-Wave business model

In my previous article on D-Wave (Oct 2024) I presented a mathematical model of the likely financial future of D-Wave, the key line items from the model are shown below; however, the reported performance of D-Wave and the changes it has made to its business mean my model no longer represents a likely trajectory. D-Wave has changed so much that I must start the model anew when we get the next set of audited accounts and earnings call.

Mathematical Model (Author)

Recent changes are so significant that it prompted me to write this update article.

D-Wave will have far less dilution than expected

The model returned a fair value of $24 and gave a warning about D-Wave's financial position, suggesting that they would need to raise $250 million to achieve positive cash flow. At the time, the share price was under $1.10 and the model predicted dilution of 150%.

On Dec 12, D-Wave announced they had completed a $75 million ATM at a share price of $4.81. The share price change has reduced the predicted dilution to 50%. The CEO confirmed in our discussion that they do not yet have all the money they need. This alone would triple my initial price target, but it is perhaps the least impactful of the changes I am referring to.

The Business Model has Changed.

D-Wave has made unexpected and significant changes to its business. Firstly, a premium pricing model. I discussed the service level agreements (SLA) in my previous article, but I now have additional information that changes my forecast of D-Wave margins. The CEO explained the SLA in some detail; it is offered to companies deploying solutions; he mentioned NTT Docomo and Pattison Foods by name. D-Wave guarantees these customers a particular level of service on their Leap Cloud QCAAS for a premium price. Those customers who are still in the research and development stage do not pay the premium.

It will make a real difference; companies with deployed solutions will be the heaviest users, Pattison relies on D-Wave to schedule their home delivery service and work out staffing schedules throughout their organization. NTT Docomo uses D-Wave to determine which cell towers need to send a "Ping" to a cellular device. It has been shown to reduce network traffic by 15%. I covered both of these use cases in my previous article about D-Waves commercial traction.

I do not have any figures about how much premium customers are paying for the SLA, but it will result in higher margins, and the impact will grow over the long term as more companies deploy their solutions.

Secondly, D-Wave has added an entirely new business that was not part of my previous analysis and is already responsible for doubling my revenue forecast for 2025.

Q4 bookings for 2024 came in at $18 million, a 500% increase on 2023, I was expecting $3.2 million. I had forecast 2025 revenue of $12 million; it is now likely to be double that figure. It is an extraordinary achievement but comes down to one key addition to the business model.

D-Wave has sold a quantum annealing computer. I tried to push the CEO to reveal as much detail as possible, but he was quite cagey. The customer and D-Wave are planning a joint statement that could be a catalyst for another share price jump soon.

From our discussion, I was able to come to the following conclusions. The new machine will likely be for the new Advantage 2 architecture and will be delivered in 2025. The machine is not the one announced for Davidson technology. D-Wave has still not revealed how the Davison machine was being financed despite me asking the question.

The name of the customer will be critical. If it is a university or government agency, it will likely reflect ongoing research and perhaps not signal a major change to long-term revenue. However, if it is going to a large commercial enterprise, then that would be a game changer, every competitive business would be forced to look at D-Waves annealer and several more orders for annealers could arrive.

Based on our discussion, I did not get a definitive answer, I suspect it is going into a commercial environment. D-Wave does not give forward guidance yet, so I could not get any information about the likely future sales of machines. It was clear that the CEO believes this will be a significant part of the company in the future.

Estimating the revenue from this new channel is very challenging. I was expecting +$3 million of bookings for Q4, including the machine, it came in at $18 million. I will assume they beat my forecast of $3 million on the existing business, implying the machine has a price of near $14 million. I don't know how that breaks down between hardware and support and have no idea about the margins involved.

The sale of 1 machine this year, 1 in 2026, and 2 in 2027 and 2028 would double my medium-term revenue targets for D-Wave and deliver a significant appreciation in the share price.

Conclusion

D-Wave will likely be the only quantum computing company with a commercially useful device for the next decade, or perhaps even longer.

The annealing technology is growing in popularity, and several high-profile companies are using it to solve difficult optimization problems.

A change in share price allowed D-Wave to raise cash by selling far fewer shares than expected, reducing the total amount of dilution expected.

Introducing service-level agreements for customers with deployed solutions will increase margins going forward.

D-Wave sold its first machine and announced record bookings last quarter. Adding computer sales likely doubles the medium-term revenue outlook for D-Wave.

In the last three months, D-Wave has improved its business position in a material way that warranted this additional article, and I reiterate my strong buy rating.

If you follow quantum stocks you will be aware of the extreme price swings we have seen recently. D-Wave stock soared from $1.10 at the time of my last article in October to over $11 in December, a 900% increase in 2 months, followed by a 56% fall to $4.80 in January. All of the Quantum stocks have moved together, suggesting these price changes are not about the fundamentals of the business but the hype surrounding the sector. Until the market realizes that D-Wave is unlike the rest, the competition is decades away from a commercial product. In contrast, D-Wave is already there. The D-Wave share price will likely continue to move in step with the other quantum start-ups. The link will eventually be broken, and once it is, D-Wave should outperform.

I am not skilled at playing these types of moves, so I will tread carefully before opening a new position. I will write in the comments section of this article when I take a position and will wait for a little calm to take some risk off the table.