Pasqal’s commercial expansion: factories, cloud access and the path to revenue

Pasqal’s delivery cycle meets its revenue model

Pasqal’s next expansion links factory capacity with the commercial work needed to sell quantum computing. On 24 September 2026, the French processor maker had seven machines operating across customer installations and internal research programmes. Annual manufacturing capacity stood at thirteen machines across facilities near Paris and in Canada. Funding following its Nasdaq listing would support commercial and technical-sales teams, technology development and additional capacity in 2027 and 2028. The company sells complete machines to businesses and computing centres, alongside time-based access through cloud channels. These routes produce revenue on different schedules. A machine contract can take more than twelve months to reach delivery; cloud revenue is recognised when customers use the platform. Consequently, an order, an installed processor and paid computing time represent different stages of the business. The same distinction applies to Pasqal’s €70.4 million of booked and awarded business: that measure combines multiyear customer contracts with grants and tax credits. It therefore reaches beyond the sales pipeline for machines alone. Reuters’ September report sets out those delivery and recognition arrangements.

Cash at two dates, losses over six months

The half-year results put dates around the expanded balance sheet. Cash and cash equivalents were €110.8 million on 30 June 2026. The subsequently announced €312.9 million balance related to 27 August, following the business combination with Bleichroeder Acquisition Corp. II and financing. Revenue for the first half was €4.872 million, compared with €4.286 million a year earlier; QPU-related services contributed €3.9 million, rising 34%. Over the same reporting period, operating cash use reached €25.194 million, versus €19.956 million in the preceding year. The operating loss was €59.161 million, while the net loss was €53.236 million. These are separate accounting measures. The company identified €27.3 million of share-based compensation and €10.2 million of one-off transaction costs, together €37.5 million, within the half-year charges. Its enlarged cash position thus followed financing after the reporting period, while the revenue and expense comparisons covered January through June. Keeping those periods visible preserves the distinction between financial resources and current commercial activity.

Building machines in Québec

The Canadian factory formed part of that production footprint before the listing. Pasqal opened its first North American quantum-computer manufacturing facility in Sherbrooke, Québec, in June 2025. It was the company’s second factory worldwide, following its French operation. The plant sits inside Espace Quantique 1, a 50,000-square-foot building; that area describes the building containing the operation. At the same time, Pasqal finalised the sale of a 100-qubit processor to Distriq, the local quantum innovation zone. Manufacturing and installation of that system were to take place in Sherbrooke. The arrangement joined a local production base to a local customer, with access intended for researchers and industry. Investissement Québec had supported the factory project through a loan announced in 2023. Funding for Distriq’s purchase involved Québec’s economy ministry, Investissement Québec, the National Bank of Canada and Canada Economic Development for Québec Regions. Distriq’s intended application areas included healthcare, energy, logistics, advanced materials and finance.

A procurement contract for Italy

A European public-computing purchase followed a different institutional route. On 27 March 2025, the EuroHPC Joint Undertaking announced the procurement contract for a Pasqal quantum computer hosted and operated by CINECA in Bologna. The EuroQCS-Italy acquisition carried a €13 million price, financed equally by the Joint Undertaking and Italy’s Ministry of Universities and Research through ICSC under the national recovery plan. The first-generation machine was specified for at least 140 qubits in analogue mode, with integration into the Leonardo supercomputer environment. A hybrid analogue-digital upgrade was planned for 2027. CINECA led the consortium with Slovenia’s ARNES and Germany’s Forschungszentrum Jülich. EuroHPC ownership, CINECA’s operating role and the shared funding arrangement were distinct parts of the purchase. The system uses neutral atoms held optically, with programmable laser interactions. Its intended research areas included materials and chemistry, optimisation, logistics and machine learning.

SOL’s inauguration and the calibration stage

The Italian deployment reached a public inauguration on 11 June 2026. SOL was unveiled at the DAMA Technopole in Emilia-Romagna, alongside LISA, an upgrade to Leonardo. At that point SOL was undergoing calibration, and availability of its computing resources was expected in the autumn. The ceremony therefore preceded the access timetable stated that day. SOL retained the first-generation specification of at least 140 analogue qubits, with the hybrid analogue-digital upgrade planned for 2027. CINECA’s site would bring quantum computation into the same institutional setting as its established supercomputing infrastructure. LISA was a separate investment: the new AI partition had 166 servers carrying 1,328 GPUs, with €50 million covering infrastructure, operation and support. Its resources were expected to become available during summer 2026. The two projects shared the inauguration and computing-centre setting, while their hardware, budgets and expected availability dates differed. SOL’s procurement, installation, calibration and subsequent user access were successive parts of the quantum deployment, each with its own practical milestone.

Aramco’s on-site installation

An industrial installation placed a processor inside an energy company’s infrastructure. In November 2025, Aramco and Pasqal announced the installation of a 200-qubit system at Aramco’s data centre in Dhahran. The machine used two-dimensional arrays of neutral atoms. The partners intended to explore applications in the energy sector, including materials work, as part of Aramco’s technology programme. Their collaboration also included plans for training Saudi engineers and scientists and supporting local research capabilities. These activities extended the relationship beyond a single equipment delivery to people able to develop and test applications on the system. Aramco’s venture investment programme, Wa’ed Ventures, had invested in Pasqal in January 2023, preceding the installation announcement by almost three years. The investment and the later machine deployment occupied different stages of that relationship. The November release described the equipment’s installation and intended industrial uses; those uses remained the partners’ development objectives. Placing a processor in a customer data centre gave that work a dedicated physical location, alongside the training and research programme envisaged for local teams.

Three cloud routes to the hardware

Azure Quantum

Microsoft Azure Quantum provided one route for organisations without an on-site Pasqal machine. An integration announced on 13 March 2025 extended access to the company’s neutral-atom technology through Microsoft’s cloud service. Its scope was analogue quantum computation. Enterprise research and development teams could explore the hardware through cloud access, avoiding the initial infrastructure commitment involved in purchasing their own system. Pasqal presented the integration as a way to broaden the group of users able to experiment with its processors. Microsoft’s service supplied a common cloud environment through which users could encounter quantum hardware, software and solutions. Within that environment, Pasqal’s contribution remained its particular neutral-atom technology and analogue capabilities. The announcement connected access to an existing enterprise cloud channel with the research work required to explore potential applications. Company claims about speed, scalability and energy efficiency accompanied the launch, but the commercial arrangement itself concerned access to computing resources.

Google Cloud Marketplace

The Google route added a marketplace purchasing channel on 13 May 2025. Customers could obtain access to a 100-qubit Pasqal processor through Google Cloud Marketplace, using a pay-as-you-go model. The computing environment was Pasqal Cloud, a separate service reached through the marketplace arrangement. Its tools allowed users to run and monitor jobs, optimise their work and post-process results. Pasqal also offered open-source libraries spanning higher-level and lower-level programming. That combination placed purchasing, job execution and development tools within the announced access route. Further high-level libraries for graph machine learning, optimisation and simulation were planned for release during 2025. Their presence in the May roadmap expressed the intended development schedule at that date. The marketplace announcement consequently had two components: an available route to the processor and a programme for extending software around it. For customers, the announced charging model tied access to use, while the service supplied the tools for submitting and following computing work.

OVHcloud’s platform

OVHcloud added another distribution setting in November 2025. Pasqal’s Orion Beta, a 100-qubit processor, became the first QPU available on OVHcloud’s quantum platform at launch. The platform had previously offered classical quantum emulation: OVHcloud listed nine emulators and approximately 1,000 users, with its emulator activity dating to 2022. A physical processor added another resource to that environment. The emulator user count described the existing platform community rather than a disclosed number of customers purchasing time on Pasqal hardware. OVHcloud also planned eight additional QPUs by the end of 2027, including seven from European suppliers. Those additions formed a future platform target. The Pasqal launch brought neutral-atom hardware into a service where users could already develop and test quantum approaches with emulators. It broadened access alongside the Azure and Google arrangements, with a separate provider and platform history.

Connecting quantum processors with supercomputers

Public research infrastructure supplied another route from installed equipment to usable workloads. Ruby in France and Jade in Germany were jointly inaugurated in November 2025 within the HPCQS project. Pasqal manufactured the processors before delivery and assembly at CEA’s Très Grand Centre de Calcul and Forschungszentrum Jülich. Each could control more than 100 qubits. Their connection with the Joliot-Curie and JURECA computing environments supported hybrid quantum-classical work. The project’s software stack combined industrial and open-source components, including Eviden Qaptiva, ParTec’s ParaStation Modulo, Slurm and Pasqal’s software development kit. Standard resource management allowed researchers to submit hybrid work through familiar computing-centre procedures. Interoperability with Qaptiva and myQLM formed part of the integration. Two use cases were presented at the inauguration to demonstrate the connection between the quantum processors and supercomputing resources. The project placed emphasis on applications such as optimisation, simulation and machine learning. For computing-centre users, the resulting infrastructure joined physical quantum hardware to scheduling, software and established classical resources, extending earlier experimentation with emulators to work on real processors.

Runtime and scheduling integration

Pasqal extended that software approach in March 2026 with an integration between NVIDIA CUDA-Q and the Quantum Resource Management Interface, or QRMI. CUDA-Q provides an open-source programming environment for combining classical GPU kernels with quantum routines. QRMI connects the runtime to computing-centre operations. Pasqal’s implementation covered Slurm-based scheduling, authentication, allocation of resources and monitoring across CPUs, GPUs and quantum processors. Those functions allow a hybrid job to fit into the centre’s existing operational framework. The company planned its first on-premises deployment of this integration at CINECA, alongside Leonardo. The capability was already available through Pasqal’s cloud service when the March announcement appeared. QRMI originated as an IBM initiative, with collaboration involving Pasqal, Rensselaer Polytechnic Institute and the UK’s STFC Hartree Centre. The integration therefore addressed the software path between a programmer’s hybrid workload and the infrastructure assigned to run it. Its announced scope concerned execution and resource management, supplying another operational layer around the processor rather than a new measure of physical-qubit capacity.

A bank’s route toward production use

Crédit Agricole CIB’s June 2026 agreement made the application-development timetable explicit. Collaboration with Pasqal began in 2019, followed by major projects from 2021 on counterparty credit-default risk and portfolio optimisation. The expanded partnership aimed at initial production applications as early as 2028, including work on risk-weighted assets and capital consumption. Three streams would advance together: quantum-inspired algorithms running on existing classical systems, operational tests on Pasqal’s physical processors, and integration toward larger hybrid applications. The first stream could enter an existing IT environment without quantum hardware, while the second tested that hardware directly. The programme also included access to simulators and processors and the development of trained business teams, strategic coordinators and project managers. The partners described performance gains under particular conditions and for specific use cases, preserving a narrower claim than a general advantage across banking calculations. Their production objective set a future date for bringing this work into daily processes. It connected years of joint research with testing, staffing and integration steps still required on the way to operational use.

Hardware ambitions retain their own timetable

Pasqal’s June 2025 roadmap separated deployment, quantum advantage and fault-tolerant computing into three development priorities. It targeted 1,000 physical qubits by the end of 2025 and 10,000 by 2028. Logical-qubit targets followed another sequence: two in 2025, twenty in 2027, one hundred in 2029 and two hundred in 2030. Logical qubits combine physical qubits to reduce errors and support longer computations. The product sequence envisaged Orion Gamma with more than 140 physical qubits by the end of 2025, Vela in 2027, Centaurus in 2028 and Lyra in 2029. Further improvements concerned repetition rates, fidelity and parallel gate operations, alongside the number of qubits. Photonic integrated circuits, following the acquisition of Canada’s Aeponyx, were intended to improve control, stability and scalability. These were dated company targets describing several technical dimensions of the planned systems. The roadmap joined those ambitions to a modular, upgradable platform, with analogue workloads providing the starting point and fault-tolerant digital computation forming the longer-term objective.

Milestones across the commercial programme

• 27 March 2025: procurement contract announced for the Italian system.
• 13 May 2025: Google Cloud Marketplace access announced.
• 11 June 2026: SOL inaugurated while calibration continued.
• 2027: planned hybrid analogue-digital upgrade for the Italian system.
• As early as 2028: targeted first production applications in the banking partnership.

Leave a comment

NameEmailWebsiteCommentSave comment