QuantWare has announced a major scaling breakthrough with the debut of VIO-40K™, a new generation of its VIO™ 3D scaling architecture for superconducting Quantum Processor Units (QPUs). The technology is designed to enable the creation of QPUs with 10 KiloQubits (10,000 qubits), which the company states is 100× larger than anything currently available in the industry.
The current quantum industry has faced scaling bottlenecks in hardware, limiting maximum QPU sizes to around 100 qubits per chip. QuantWare’s VIO 3D architecture solves this issue by delivering signals directly to qubits in a three-dimensional stack of chiplet modules connected via ultra-high-fidelity chip-to-chip connections. This approach removes the exponential signal routing that consumes space on 2D chips, allowing the new architecture to provide exponentially more compute per dollar and per watt compared to systems that rely on networking many smaller processors.
The VIO-40K architecture is designed to become the industry scaling standard under the Quantum Open Architecture (QOA) framework, meaning it is capable of scaling up any superconducting qubit design. Furthermore, QuantWare announced that NVIDIA NVQLink, the open platform for building logical QPUs, is part of the VIO-40K compatible ecosystem. This integration will provide a low-latency, high-throughput connection to classical AI supercomputing via the NVIDIA CUDA-Q platform, enabling the tight coupling required for fault-tolerant quantum computing.
To meet expected demand for the architecture, QuantWare announced it is building Kilofab, an industrial-scale QPU fab scheduled to open in Delft, The Netherlands, in 2026. Kilofab will be the world’s first dedicated fab for QOA devices and is expected to increase QuantWare’s production capacity by 20×. The first VIO-40K devices are slated for customer delivery in 2028.
Read the full announcement here.
December 10, 2025
Leave A Comment