Quantum computing has entered a new phase. Instead of chasing the long-term ‘qubit race’ to surpass classical von Neumann architectures, the field now prioritizes tighter integration with classical infrastructure. Inside leading supercomputing centers, teams are pairing quantum processors with GPU- and CPU-based systems to tackle specific workloads, while governments worldwide intensify support for quantum hardware and supply chains..
Quantum’s allure remains its uniquely nonclassical properties – superposition and entanglement – which promise new ways to represent and manipulate high-dimensional correlations beyond binary 0/1 machines. The focus is now on making those properties usable alongside existing data center infrastructure..
Policy Tailwinds for Quantum Infrastructure. In May this year, the Department of Commerce announced more than $2 billion in incentives to accelerate quantum commercialization, including support for quantum manufacturing and the development of utility-scale, fault-tolerant systems.
In June, the White House issued an Executive Order on the “Next Frontier of Quantum Innovation,” intended to shift near-term attention toward practical quantum components, supply chain procurement, and infrastructure security. Other governments have advanced similar initiatives, signaling that quantum-classical integration is a multiyear priority..
Related:FLOPS vs Megawatts: Who’s Winning in 2026 Supercomputing?. Modalities and the Integration Challenge. Quantum hardware spans several modalities – superconducting, trapped-ion, neutral-atom, photonic, and silicon spin, among others – each with distinct environmental and operational requirements.
Superconducting systems, for example, rely on deep cryogenic cooling and are sensitive to thermal, magnetic, and vibrational interference. Because modalities are evolving rapidly, deployments must accommodate add-ons and refurbishments without major disruptions to adjacent classical infrastructure..
Growing public and private investment is reordering priorities. Rather than chasing demonstrations of “quantum supremacy” – single tasks that exceed classical capabilities – the industry is building hybrid quantum-classical architectures that deliver measurable gains on real workloads.
While physical scaling and infrastructure hurdles remain formidable, operators are colocating quantum processing units (QPUs) with GPU/CPU nodes where low-latency coupling can reduce time-to-solution.. From Qubit Counts to Hybrid Performance. “The qubit count issue is in some sense becoming much less important because you’re now addressing end users who are interested not in what your particular modality or your technical specifications are,” Bob Sorensen, chief quantum computing and AI analyst at Hyperion Research, told Data Center Knowledge.
“Today, quantum computing is shifting from science experiments to productization. Organizations are thinking now about how they introduce quantum computing into our classical compute environment.”. Related:NSF’s $20M Quantum Push: What It Could Mean for Future Data Centers.
Sorensen said that Hyperion estimates continued revenue growth in quantum computing, with the market at $1.4 billion in 2025 and a 30% annual growth rate out to 2028, reaching about $3 billion. The most promising near-term applications center on quantum-level simulations, particularly in computational chemistry and materials science.
By Hyperion’s assessment, digital simulators based on CPUs and GPUs account for almost one-quarter of the quantum computing hardware market the firm tracks, with GPUs outpacing CPUs by roughly 2x in this segment.. Ecosystem Pivots to Hybrid. Vendors are aligning around hybrid roadmaps:.
In March 2026, IBM outlined methods for “quantum-centric supercomputing,” a staged approach that starts with offloading specific calculations to quantum systems and progresses to co-designed, heterogeneous systems built from the ground up.. At GTC 2026, Nvidia introduced its NVQLink architecture to connect QPUs to GPU supercomputers and announced a collaboration with Quantum Machines on an open framework that integrates classical systems and Nvidia GPUs within quantum control stacks..
In June 2026, Hewlett Packard Enterprise said it is working with Intel, IQM, Qblox, Quantinuum, QuEra Computing, Quantum Machines, Rigetti, and Riverlane on algorithm co-design and software interoperability to connect different styles of quantum computers to its Cray platform in hybrid configurations.. Also in June, AMD said it is working with OQC and JPMorgan Chase to explore how quantum computing, AI, and high-performance classical infrastructure can address complex financial services workloads..
Related:Quantum Progress Runs Through the Data Center – AWS Shows Why. Inside the Data Center: Tighter Quantum-Classical Coupling. At the Quantum.Tech World conference, held in Boston in June, quantum-classical integration was front and center.
Henning Soller, a partner at McKinsey, told attendees that early integration work is needed now, ahead of possible clarity on a quantum computing advantage in the 2028-2030 timeframe.. “One of the key aspects of making quantum computing usable is not just developing the quantum computers, but also developing the integrations with the high-performance computers with classical infrastructure,” Soller said, noting that the bulk of the data will continue to reside in conventional data center databases..
McKinsey’s research indicates that colocating quantum systems with classical infrastructure can improve overall hybrid performance by reducing communication latency for certain workflows.. “One of the key aspects of making quantum computing usable is not just developing the quantum computers, but also developing the integrations with the high-performance computers with classical infrastructure.” – Henning Soller, partner at McKinsey.
Facility Impacts: Power, Cooling, and Vibration Control. On the same Quantum.Tech stage, Aparna Prabhakar, chief strategy and sustainability officer for Schneider Electric’s energy management business, urged operators to take power and cooling requirements seriously as quantum moves from lab to production.
She pointed to the unfolding story of generative AI in data centers, where retrofitting is underway as the electricity grid tries to keep up with AI innovation. Quantum-classical hybrid deployments likewise merit full, system-level consideration of cooling, processing, and power..
Prabhakar noted that teams should evaluate whether cooling approaches entail changes at the slab level and how to mitigate vibration. “You need a floor setup that is not going to disturb the quantum computer,” she said.. Algorithms and Access.
Also at Quantum.Tech this year: a snaking line for selfies that stretched the exhibit hall, leading to none other than Peter Shor. Attendees queued to meet the gray-haired MIT professor and mathematician behind Shor’s algorithm, which, in 1994, showed that a quantum computer could factor ultra-large numbers – something well beyond what classical von Neumann supercomputers could do.
The algorithm became an essential reference point in fault-tolerant quantum computing.. Peter Shor (center) speaking at the Quantum.Tech World conference in June 2026. (Image: Quantum.Tech World). Asked why relatively few quantum algorithms have emerged since, Shor said: “We haven’t discovered very many more algorithms … and the ones we have discovered are for abstruse problems that nobody actually wants to solve in practice,” with some exceptions in physics and quantum chemistry simulations.
He suggested that many classical algorithms would never have been discovered without broad experimentation on classical computers. The implication is that greater hands-on access to quantum systems may catalyze new algorithmic breakthroughs. That hope mirrors the ambitions of those now blending quantum and classical capabilities into a co-dependent ecosystem.
