IN Brief:
- D-Wave has delivered a 4,400-plus-qubit Advantage2 system to Villa Saporiti in Como, Italy, for installation.
- Swiss Quantum Technology holds rights to 50% of the system's capacity for five years under a €10m agreement.
- The computer is intended to become accessible through D-Wave's Leap cloud service after installation and commissioning.
D-Wave Quantum has delivered an Advantage2 annealing quantum computer to Lombardy, Italy, ahead of its installation at Villa Saporiti in Como. The system contains more than 4,400 superconducting qubits and will operate within the Q-Alliance initiative under the name VoltaB1, providing a physical computing platform intended for European research and commercial applications.
The arrival was announced on 15 October at Digital Innovation Forum 2026. It follows a €10m agreement signed with Swiss Quantum Technology SA in October 2025 to deploy an annealing quantum computer in Europe. Under that agreement, the Swiss company acquired rights to 50% of the system’s capacity for five years and holds an option to purchase the complete machine.
VoltaB1 has arrived for installation rather than entering confirmed routine operation. D-Wave plans to connect it to the company’s Leap quantum cloud service once the system is operational, allowing remote users to submit computational problems without visiting the facility. The announcement does not provide a completed commissioning record, an operational start date or performance results measured at the Como installation.
The machine was scheduled for formal inauguration at Q-Alliance’s Quantum per l’Italia event on 16 October. That ceremony is distinct from the engineering process required to install and validate the hardware. A functioning quantum computing installation must meet electrical, environmental, control and networking requirements before applications can be executed under the intended operating conditions.
Advantage2 is D-Wave’s sixth-generation quantum annealing platform. Its computing approach differs from that of universal gate-model machines, which perform sequences of quantum logic operations. Quantum annealers are designed to search for low-energy configurations of physical systems representing particular mathematical problems, making them relevant to certain optimisation and simulation workloads.
Many optimisation problems can be expressed through variables and mathematical relationships that describe desirable or undesirable combinations of values. In an Ising formulation, interacting binary variables contribute to an energy function, and the computational objective is to identify a configuration with a low energy value. A related quadratic unconstrained binary optimisation formulation represents the problem using binary variables and pairwise terms.
D-Wave’s processor allows selected coefficients in that energy function to be programmed into the hardware. The quantum system then undergoes an annealing process intended to produce candidate low-energy configurations. Results are measured and returned to conventional computing resources for evaluation or further processing. Multiple runs may be required to assess the quality and consistency of the solutions obtained.
The approach depends on whether the original application can be represented effectively using the available hardware structure. Practical problems often contain constraints, variables and relationships that must be reformulated before submission. The resulting mapping can require additional physical qubits, particularly when logical variables need connections that are not directly available between neighbouring elements of the processor.
Advantage2 uses D-Wave’s Zephyr connectivity architecture, providing up to 20 connections per qubit compared with the 15-way connectivity of the preceding Pegasus topology. Greater connectivity can reduce the additional hardware resources required to represent certain problems, although the benefit depends on their structure and the embedding method selected.
The sixth-generation processor contains more than 4,400 qubits and over 40,000 programmable couplers. These couplers provide the interactions through which the relationships between variables are represented. The physical qubit count should not be interpreted as an unrestricted number of application variables because the hardware connectivity, calibration and problem formulation determine how many physical resources are needed for a particular calculation.
The processor is built using superconducting circuits, requiring operation at extremely low temperatures. D-Wave’s technical documentation identifies a dilution refrigerator that maintains the quantum processing unit below approximately 20 millikelvin, together with radio-frequency and magnetic shielding intended to reduce external interference. Conventional electronics remain necessary to configure the system, communicate with its processor and manage measurement results.
These requirements make the Como deployment substantially different from installing conventional computer servers. The physical installation must accommodate cryogenic equipment, electrical supplies, suitable environmental conditions and the interfaces connecting the quantum processor with external control and computing resources. The manufacturer has not published a site-specific acceptance record confirming that all these requirements have been satisfied.
Advantage2 became generally available in May 2025, when D-Wave introduced the 4,400-plus-qubit platform through its Leap service. The company reported a 40% increase in energy scale, twice the coherence time and a 75% reduction in noise compared with its previous-generation Advantage system. These figures describe manufacturer-reported platform comparisons, not tests carried out following delivery to Italy.
D-Wave has also published research claiming performance advantages on selected optimisation and materials-simulation problems. The usefulness of such comparisons depends on the problem instance, accuracy requirements, classical algorithms selected and total computational work involved. Results obtained under one benchmark do not establish that quantum annealing is faster or more accurate than conventional computing for every optimisation application.
The Leap service combines access to quantum processing hardware with software tools and hybrid solvers that can use conventional computing alongside annealing. D-Wave advertises 99.9% availability and uptime for the broader service. That figure concerns the existing cloud offering and cannot be used to establish the operational availability of VoltaB1 before the Italian installation has been commissioned.
Q-Alliance is developing the Como facility as a centre for quantum research, application development and skills. Access to installed hardware could allow universities and businesses to test workloads against a physical annealing processor rather than relying exclusively on software simulation or remote systems outside the region. The extent of future use will depend on operating arrangements, access allocations and the suitability of particular applications for the annealing architecture.
The project follows the Volta Declaration signed in Milan in June 2026, which established a framework for cooperation involving quantum technology, artificial intelligence and computing infrastructure. The arrival of Advantage2 provides a defined hardware element within that programme, while the broader research and industrial commitments require subsequent development.
The immediate milestone is the delivery of an established quantum computing platform under the 2025 commercial agreement. Installation, commissioning, the beginning of operational access and evidence from applications executed on VoltaB1 remain separate developments. D-Wave has not announced a completed customer workload or independently verified performance result from the new Como installation.
