QuantWare has unveiled what it describes as the quantum hardware industry’s most significant scaling breakthrough to date, announcing VIO-40K, a new generation of its VIO quantum processor architecture capable of supporting QPUs with up to 10,000 qubits.
The milestone represents a two-order-of-magnitude jump over today’s largest commercially accessible processors, marking a potential inflection point in quantum hardware scaling after nearly a decade of incremental progress.
While major industry players such as Google, IBM, and several well-funded startups have advanced quantum processor designs, the physical size and wiring complexity of superconducting qubit systems have capped practical device sizes at roughly 100 qubits. Google moved from a 53-qubit to a 105-qubit chip over six years; IBM recently introduced a 120-qubit system and projects similar scales in its public roadmap through 2028. The challenge has not been conceptual but architectural: routing thousands of high-fidelity control lines to cryogenic chips without degrading qubit performance has proven prohibitively difficult.
QuantWare argues that its VIO 3D scaling architecture fundamentally removes this bottleneck. Instead of relying on monolithic chip designs or traditional packaging, VIO-40K organizes processors into chiplet-based modules connected through ultra-high-fidelity chip-to-chip links. The platform supports up to 40,000 I/O lines – enough, the company says, to enable 10,000 physical qubits within a package smaller than today’s ~100-qubit processors. Unlike networked multi-chip quantum systems, which suffer from latency and fidelity degradation across interconnects, VIO-40K is designed to operate as a single cohesive processor. QuantWare positions this as a direct route to drastically lower total cost of ownership by reducing the need for complex routing, extensive networking, and multi-processor system management.
Because VIO is offered as an open scaling standard, QuantWare is positioning itself not only as a hardware provider but as a platform enabler for the broader superconducting quantum ecosystem. According to the company, any organization that currently designs its own superconducting qubits – from startups to national labs – could adopt VIO to build larger-scale devices without redesigning core qubit technology.
QuantWare is also expanding its ecosystem around VIO through its Quantum Open Architecture initiative. NVIDIA’s NVQLink, the company’s platform for logical QPU development and integration with classical AI supercomputing, has been added as a VIO-40K–compatible component. Combined with VIO, NVQLink enables tight coupling between hyper-scaled quantum processors and NVIDIA’s CUDA-Q toolchain, allowing quantum developers to leverage GPUs and QPUs in unified workflows. Given the growing shift toward hybrid quantum-classical architectures, the linkage between quantum hardware and hyperscale AI compute may become increasingly strategic.
To support volume manufacturing of VIO-40K devices, QuantWare announced plans for Kilofab, an industrial-scale fabrication facility scheduled to open in 2026. Positioned as the world’s first dedicated fab for Quantum Open Architecture devices, Kilofab is intended to increase the company’s production capacity 20-fold. It will be located at QuantWare’s headquarters in Delft, where the company originated as a spin-out from the QuTech research institute. QuantWare reports customers in more than 20 countries and identifies itself as the largest commercial quantum hardware provider by shipment volume.
QuantWare CEO Matt Rijlaarsdam said the company believes the quantum industry has been constrained more by hardware scaling limitations than by algorithmic potential. “For years, people have heard about quantum computing’s potential to transform fields from chemistry to materials to energy, but the industry has been stuck at 100-qubit QPUs,” he said. “QuantWare’s VIO finally removes this scaling barrier, paving the way for economically relevant quantum computers.”
VIO-40K devices are expected to ship to customers beginning in 2028, and reservations are now open.
Executive Insights FAQ
Why is VIO-40K seen as a breakthrough for quantum scaling?
It removes long-standing wiring and packaging limitations by using chiplet-based modular architecture capable of routing 40,000 I/O lines, enabling 10,000-qubit processors in a single package.
How does VIO differ from quantum networking approaches used elsewhere?
Instead of connecting multiple small processors over low-fidelity links, VIO creates high-fidelity chip-to-chip connections that allow large QPUs to function as unified devices.
What role does NVIDIA’s NVQLink play in the ecosystem?
It provides fast, low-latency integration between quantum processors and classical AI supercomputing, supporting hybrid compute models through the CUDA-Q environment.
Why is the Kilofab facility significant?
It will become one of the world’s largest dedicated quantum fabrication hubs and scale QuantWare’s output to meet demand for large quantum processors.
When will customers be able to deploy VIO-40K hardware?
Reservations are open now, with first commercial deliveries targeted for 2028.
