Quantum computers edge toward industrialization - InformationWeek

IT InfrastructureIT InfrastructureDigital TransformationIT StrategyIndustry TrendsQuantum computers edge toward industrializationRecent investments in the U.S. and Europe aim to boost production, but the industry still faces scalability, error correction and software challenges.John Moore,Contributor,Informa TechTargetJune 4, 20269 Min ReadGetty ImagesQuantum technology is edging closer to industrialization as national governments commit billions to scale the technology, while hardware and software providers — sometimes with the support of enterprise customers — push toward useful systems. In May, the U.S. federal government unveiled a program earmarking $2 billion in planned funding to nine quantum computing companies. The U.K. government in March launched a £2 billion quantum innovation effort to rollout quantum computers at scale. Other countries are also investing in their national quantum ecosystems. The moves signal a push to make quantum a practical computing tool, with quantum processing units eventually taking their place in data centers alongside CPUs and GPUs. But while quantum's arrival might seem inevitable, the technology still faces significant obstacles. The Department of Commerce announcement outlining the U.S. government's quantum investment strategy cited "unresolved engineering problems." Industry executives, meanwhile, cited manufacturability, technical hurdles such as quantum error correction (QEC) and software as areas of focus. Related:Oak Ridge Puts Quantum Supercomputer Integration to the TestDavid Mooter, principal analyst at Forrester, noted that uncertainty remains, even as quantum technology appears poised for industrialization. However, he's become more optimistic about quantum technology over the past couple of years. He pointed to the Electronic Numerical Integrator and Computer (ENIAC), the first programmable electronic computer, built in the 1940s, as an analogy for quantum's current state. "We're still in the pre-ENIAC phase; we're still a ways off," Mooter said. "But I feel like the endgame is in sight, because we are now in the engineering phase rather than just the scientific theorizing phase." Boosting manufacturability Making quantum easier and cheaper to manufacture is on the minds of quantum technologists worldwide. Boosting production capacity is central to building out quantum as an industry as opposed to a science experiment. Consistent and reliable chip production supports commercial viability. The recent U.S. investment plan appears heavily weighted to industrialization. Nearly 70% of the money pledged is for establishing domestic quantum foundries: The government said GlobalFoundries will receive $375 million in planned funding, while IBM will get $1 billion. Those investments cover several modalities: GlobalFoundries will pursue superconducting, trapped ion, photonic, topological and silicon spin technology, while IBM will focus on superconducting wafers. Related:InformationWeek Podcast: Is quantum computing slumbering?Such industrialization plans aim to tap existing manufacturing approaches — namely, those the semiconductor industry developed over decades. Both IBM and GlobalFoundries have long histories in microelectronics. IBM plans to launch Anderon, a company that will serve as its quantum foundry, on the back of its experience in fabrication tools. GlobalFoundries, an AMD chip spinoff founded in 2009, said in its investment announcement that quantum companies and other innovators can build upon its established "industrial layer." This manufacturing strategy is also prevalent at companies such as Quobly, a French company that focuses on silicon qubits. Quobly partners with semiconductor maker STMicroelectronics. Maud Vinet, Quobly's CEO and co-founder, described the manufacturing challenge as moving from the relatively small number of qubits of today's quantum prototypes to fabricating millions of qubits. Applying semiconductor economics to quantumThe key to making that happen is "scaling down," she said. That is, the ability to reuse the same principles that have driven the semiconductor industry: increasing integration density and reducing the size of components. The manufacturing know-how that packs millions of transistors into cell phones can be transferred to ramp up qubit counts. Related:Backplane to the Future: InfiniBand Technology Meets Quantum"This applies not only to qubits, but also to control electronics and interconnects," Vinet said. "By integrating these elements more tightly, we can reduce system complexity and footprint." Integration lowers cost over time. But, more importantly, it also enables scalability, which Vinet called the "central challenge in quantum computing." Addressing error correction and the broader scalability issue According to Mooter, scaling is the main problem facing the quantum technology industry. Within scaling, error correction is the biggest barrier, he noted. "One is linked to the other," Mooter said. "'How do we solve those?'" is the main focus right now.QEC aims to address the fragile nature of quantum states. States such as superposition and entanglement enable quantum computation, but they easily collapse due to vibration, radiation or other disturbances. Quantum "noise" introduces computation errors, and that becomes a barrier when scaling systems. "As quantum computers scale, errors increase quickly," said Marco Ghibaudi, vice president of engineering at Riverlane, a Cambridge, U.K.-based company that provides QEC technology. "Without quantum error correction, those errors accumulate too fast for useful computations to run." He noted that error correction is best understood as a systems issue."What makes it challenging is that it sits at the intersection of hardware, control systems, classical processing and software," he said. "It is not something that can be solved in isolation." QEC is the key step in turning quantum computing into a practical technology, Ghibaudi said. He noted that a 2025 Riverlane survey found 95% of the more than 300 quantum professionals polled view QEC as essential for reaching utility-scale quantum computing. Riverlane defines utility-scale as the ability for a quantum computer to perform 1 trillion reliable operations. This is projected to happen from 2033 onward, according to the company's QEC technology roadmap, which was published in March. At the utility-scale stage, quantum machines are expected to provide "transformative advantages" in areas such as molecular chemistry, drug design and climate modeling, according to Riverlane. Cooling as a differentiatorOther issues in scalability include manufacturability, as noted, and cooling demands. As for the latter, some modalities can only operate in extreme cold. Superconducting qubits, for example, must be chilled to near absolute zero. Cooling requirements increase as the number of qubits in superconducting quantum computers expands, Mooter said. The cryogenic systems supporting superconducting quantum computers require considerable space and energy to operate. "The cooling problem is going to create a barrier to scaling," he said. That issue could result in modalities leapfrogging each other over time. Superconducting qubits currently have momentum, since they are created in manufacturing fabs similar to those used to make silicon chips, Mooter said. That makes them easier to produce than other modalities, he added. But due to the cooling challenge, neutral atom qubits could supersede superconducting systems toward the end of the 2030s, Mooter predicted. Neutral atoms are easier to chill with much less power and equipment.Building the software stackA quantum computer's software provides the problem-solving layer for enterprise challenges such as simulation and optimization. But that technology stack is very much a work in progress. Vinet said it will take some 10 to 15 years before quantum computing reaches the stage AI is at today. "Reaching a level of maturity comparable to AI will require not only advances in hardware but the development of robust software ecosystems and clearly identified applications," she said. Catching up requires progress across a number of areas. One tricky task is finding ways to efficiently load classical data into a quantum system. A quantum system can't directly import data the way a conventional system can, so, data must instead be encoded into a quantum state. That process, referred to as state preparation, carries a complexity cost: the encoding task, in the worst case, scales exponentially with the number of qubits used. "State preparation has quietly been the binding constraint on many promising quantum finance algorithms," said Georgios Korpas, principal scientist at HSBC, a global banking and financial services organization.That constraint makes it difficult to realize the promise of quantum amplitude estimation, a technique that can boost the Monte Carlo simulations used in financial services and other industries to make predictions. A Monte Carlo simulation runs data samples through a model thousands, or potentially millions, of times to generate more accurate predictions. But with quantum amplitude estimation, a simulation that would require 1 million samples on a classical machine might require only 1,000 on a quantum computer to achieve the same level of accuracy. That reduction is called a quadratic speedup. This boost is particularly evident when simulations demand higher and higher accuracy. Korpas said that quantum amplitude estimation offers a quadratic speedup for Monte Carlo simulations in applications such as value-at-risk, conditional value-at-risk, derivative pricing and credit portfolio risk. "But only if we can load the input distribution into a quantum state at a cost that does not offset the speedup," he added.That's the crux of the research HSBC pursued with Haiqu, a quantum software company based in New York. Their joint research, the results of which were published in April, focused on encoding heavy-tailed probability distributions used in financial services and other industries. Heavy-tailed distributions allow for more frequently occurring outliers than a normal distribution, which makes them useful for modeling extreme financial market events. The companies used matrix product state (MPS) methods to encode the heavy-tailed distributions. MPS let researchers encode those distributions into "shallow" quantum circuits, which were executed on IBM quantum machines with up to 156 qubits. A shallow circuit avoids the exponential gate counts that would create an intractable encoding bottleneck. A shallow circuit also means the noise on a quantum device does not substantially alter the encoded state.Mykola Maksymenko, co-founder and CTO of Haiqu, said this encoding method is the initial step toward developing software capable of addressing important business challenges. "It's not solving the whole application," he said. " But it solves this first, fundamental part, which now gives real hope and a kind of expectation that we can solve this problem on a quantum computer. If you cannot model the correct, real-world distribution, it's impossible to do any of those simulations." "Practically, that means we can prepare states encoding more realistic risk-factor distributions … that better match observed market behavior," Korpas said, citing Lévy, Gamma and other heavy-tailed distributions.Moving toward useful systemsAs the march toward commercialization continues, the quantum computing field is shifting from conducting experiments to closing in on useful systems, Maksymenko said. "A few years back, most of quantum research was happening either in completely theoretical settings or on some very basic examples, 5-qubit or 10-qubit experiments on the actual quantum hardware," he said.While those efforts looked "toyish and not really serious" in their limited scope, quantum technology has now entered a different era, Maksymenko said. "It always seems like there is still some distance to go," he said. "However, we are getting very close to a moment where very niche applications … will perform better on the quantum computer than on the classical computer."Those niche applications might surface in quantum chemistry, optimization and perhaps in some financial modeling, Maksymenko noted.But quantum computing, even as it evolves, won't supplant classical computing, Forrester's Mooter said. Quantum computers harness a fundamentally different physical phenomenon than classical computers, which makes quantum systems more efficient for certain problem sets, he said. But most problem sets are better served by classical computers, he added.In Mooter's view, quantum computers will augment conventional computing, working with GPUs and CPUs. These systems will combine to accomplish compute tasks individual systems couldn't accomplish on their own. "When quantum computers become viable, the standard architecture will be hybrid," he said.Read more about:Industry PerspectivesQuantum ComputingData ManagementIT Leadership and StrategyAbout the AuthorJohn MooreContributor, Informa TechTargetJohn Moore is a freelance writer who has covered business and technology topics for 40 years. He was previously an industry editor at Informa TechTarget and, before that, wrote for publications, including Smart Partner, Federal Computer Week and Computer Systems News. He currently focuses on enterprise IT strategy, AI adoption and integration, data management, partner ecosystems and emerging technologies.See more from John MooreWant more InformationWeek stories in your Google search results?Add Us NowMore InsightsWebinarsCyber Resilience in 2026: Maintaining Business Continuity Beyond the BreachChoose Your EUC Adventure: AVD, Windows 365, and IntuneNext Gen Data Governance for the AI EraModernizing Database Migrations: From Manual to Automated to Agent-Assisted WorkflowsDecoding the 2026 Cloud Management LandscapeMore WebinarsIndustry ReportsBeyond Legacy Virtualization: Wind River's Modern Alternative for Enterprise InfrastructureThe Forrester Wave™: Unified Vulnerability Management Solutions, Q3 2025IDC MarketScape: Worldwide Exposure Management 2025 Vendor AssessmentInformationWeek 2025 US Tech and IT Salary ReportWhy IT Contracts Need Special Attention From Lawyers and TechnologistsAccess More ResearchEditor's ChoiceMachine Learning & AIThe AI infrastructure bottleneck is becoming a CIO problemThe AI infrastructure bottleneck is becoming a CIO problemResponsible AIETS CIO on competing with AI startups 'running with scissors'ETS CIO on competing with AI startups 'running with scissors'Want more InformationWeek stories in your Google search results?Virtual Event | June 11, 2026Join us for this free virtual event as we examine how the rapid adoption of public cloud GenAI services has added new dimensions of complexity to an already challenging puzzle.RegisterGet a snapshot of the issues affecting CIOs, three times a week in your inbox.Subscribe to our newsletters today.SIGN UP