Quantum computing success hinges on service model not hardware
The race for quantum supremacy has long been framed as a hardware contest. Companies compete fiercely over qubit counts and error rates, each claiming their architecture - annealing, trapped-ion, superconducting - will ultimately prevail. This debate carries a quiet assumption: that the company building the most powerful machine will naturally capture the market. Yet that assumption overlooks a crucial detail about how technology actually reaches paying customers.
History offers a useful parallel. Consider the early days of personal computing, when hardware specifications dominated headlines and marketing campaigns. The machines that ultimately won were not always those with the most impressive technical specifications on paper. Success often belonged to companies that solved the more mundane problem of making the technology accessible, useful, and worth paying for on a regular basis. The same pattern may now be repeating in quantum computing, where the software and cloud layer could matter more than raw processing power.
The industry currently sits in what analysts describe as a noisy middle period. Quantum machines can operate with up to roughly 1,000 qubits, yet errors remain frequent and persistent. Fully fault-tolerant quantum computing - systems that correct their own errors reliably - appears to be years away from practical reality. This means hardware cannot fully scale in the near to medium term, leaving a gap between what the machines promise and what they can actually deliver for commercial users.
That gap creates an opening for a different kind of business model. Rather than selling expensive hardware that clients must operate and maintain themselves, companies can offer access to quantum tools through the cloud. This approach, known as Quantum Computing as a Service or QCaaS, allows clients to run hybrid classical-quantum systems and complete vital tasks on quantum processors without ever owning the physical hardware. The model transforms a capital-intensive purchase into a recurring service relationship.
Why Service Revenue Beats Machine Sales
The distinction between selling hardware and selling access carries significant financial implications. Hardware sales tend to be lumpy - large purchases that arrive irregularly and make revenue difficult to predict. Service models, by contrast, generate steadier income streams that investors can forecast with greater confidence. For quantum companies still searching for sustainable business models, this difference matters enormously.
D-Wave Quantum Incorporated, trading on NASDAQ under the symbol QBTS, has built one of the strongest cases for the QCaaS approach. [1] The company has negotiated several eight-figure enterprise agreements for its quantum computing services this year, with particularly strong results in the most recent quarter. These substantial contracts complement the company’s hardware system sales, providing a more predictable revenue foundation alongside the irregular equipment purchases.
The service business showed particular strength even when overall revenue disappointed. In the latest quarter, D-Wave’s commercial metrics stood out as the bright spot amid otherwise lackluster financial performance. The company reported that more than 37 percent of its QCaaS revenue now comes from production applications - meaning companies are using quantum computing for live and operational workloads, not merely for experiments or research trials. [1]
This figure carries weight because it signals something important about customer behavior. Companies running production workloads on quantum systems are far more likely to become stable, repeat customers than those still testing the technology. They have moved beyond curiosity into actual dependence, creating the kind of recurring subscription revenue that service businesses need to thrive. Each production deployment represents a customer that has integrated quantum computing into its actual operations.
Three Companies, Three Different Bets
The quantum computing sector features several publicly traded companies pursuing distinctly different strategies. D-Wave has positioned itself most aggressively around the service model, using its annealing-based architecture to solve optimization problems for enterprise clients. Rigetti Computing, trading under RGTI, shows stronger revenue and margin growth but remains more dependent on hardware sales. [2] Quantum Computing Incorporated, under QUBT, sits furthest from a service-based model and carries the highest risk profile of the three. [3]
Rigetti’s approach reflects a bet on superconducting technology and the belief that hardware excellence will eventually translate into market dominance. The company has demonstrated impressive growth metrics, yet its continued reliance on equipment sales leaves it exposed to the same lumpiness that has complicated D-Wave’s financial picture. Without a robust service layer, hardware makers must constantly find new buyers for expensive machines rather than cultivating long-term relationships with existing customers.
Quantum Computing Incorporated represents the most speculative investment among the three. The company remains furthest from establishing a credible QCaaS offering, which amplifies its risk for investors. While all three firms face the fundamental challenge of an immature market, QUBT’s limited progress toward service-based revenue makes its path to profitability less clear than its competitors’.
The service layer may ultimately determine which quantum companies succeed. Hardware capabilities will continue improving, and the debate over architectures will persist. But the companies that figure out how to make quantum computing accessible, reliable, and worth paying for on an ongoing basis may capture the market regardless of which underlying technology ultimately proves most powerful. The noisy middle period, frustrating as it is for those waiting on fault tolerance, may ultimately reward the companies that build durable service relationships rather than those that merely build better boxes.se awaiting fault-tolerant machines, may actually be the moment when business models get tested and proven.