MadQCI Cloud: Quantum As A Service
2 July 2026
For a long time, quantum technologies have been confined to highly specialized laboratories, requiring massive investment and doctorate-level expertise to operate. However, the MadQCI project has successfully pioneered a “Quantum as a Service” (QaaS) model that mirrors the convenience of modern cloud computing. By deploying a distributed, data-center-like infrastructure across various metropolitan nodes have created a testbed where quantum capabilities are no longer a rare luxury but an on-demand resource available to third-party developers.
The delivery of quantum services through MadQCI relies on the innovative “anything as a service” (XaaS) paradigm, which hides the underlying technical complexity from the end user. Instead of needing to understand the nuances of photon states or cryogenics, a user interacts with the network through standardized REST APIs and well-known software interfaces. This approach allows for hardware abstraction, meaning that the physical QKD (Quantum Key Distribution) systems operate as “black boxes” in the background. The infrastructure is built on high-virtualization IT servers connected by high-speed optical fibers, mimicking the setup of a commercial cloud provider. This allows the network to provide multi-tenancy, where multiple users can share the same expensive quantum hardware simultaneously through virtualized key streams.
The range of services offered by MadQCI spans every level of the traditional cloud stack, categorized as Quantum-IaaS, PaaS, and SaaS. At the infrastructure level (IaaS), users can access virtualized quantum links and manage the raw delivery of keys across different network spans. Moving up to the platform level (PaaS), developers are provided with software development kits (SDKs) and programmable methods to build their own quantum-enabled applications, such as specialized cryptographic tools. Finally, at the software level (SaaS), MadQCI offers ready-to-run services like Entropy as a Service (EaaS). This particular service utilizes Quantum Random Number Generators (QRNG), such as the Quside Garnet, to provide high-quality, unpredictable entropy that is essential for everything from scientific simulations to secure gambling platforms.
One of the most critical applications of this infrastructure is its role in securing the future of the internet through Quantum-Safe TLS and IPsec. Current security protocols like Transport Layer Security (TLS), which protects website traffic, and IPsec, which powers Virtual Private Networks (VPNs), rely on classical key agreement methods that are vulnerable to being broken by future quantum computers. MadQCI addresses this threat by integrating quantum-distributed keys into these existing protocols. Instead of relying on mathematical problems that a quantum computer could solve, these “quantum-safe” versions use symmetric keys generated through the laws of quantum mechanics, providing what is theoretically unconditional security.
In the case of IPsec, MadQCI provides a specialized REST API that allows a user to deploy a secure, quantum-resistant tunnel on demand. When a request is made, the system automatically retrieves a 32-byte quantum key from the SD-QKD software stack and uses it to establish an encrypted tunnel between two remote nodes. This process is entirely transparent to the user, they simply send their data through the designated interface and the infrastructure ensures it is protected by quantum-distributed material. Similarly, Quantum-Safe TLS services are delivered by hybridizing traditional protocols with quantum keys, ensuring that client-server communications for websites and APIs remain resilient against both current and future threats.
The scalability of these services is a cornerstone of the MadQCI design. The infrastructure can support multiple concurrent end-user instances, up to a thousand in some configurations, by managing the “secure key rate” provided by the hardware. For example, the high-performance Toshiba systems can generate enough key material to support a vast number of TLS sessions every second across a 23-kilometer metropolitan span. By making these advanced security primitives accessible, programmable and affordable, with some TLS service instances costing only a tiny fraction of a cent in equipment depreciation, MadQCI is building the foundational architecture for a quantum-safe digital ecosystem. Through this cloud-based approach, Madrid is proving that the quantum future is not a distant dream, but a service that can be requested with a simple line of code.
Several examples of QKD systems deployed or being tested in the Madrid quantum network. The technology procured for the MadQCI ecosystem was selected to fit each application scenario, not the other way around. In this way, this technology enables other technologies specific to each application, allowing for the generation of applied and transferable knowledge, and, since each piece of technology is best suited to different uses, the MadQCI ecosystem is a showcase for the entire European quantum ecosystem, with practically all of its manufacturers represented. (a) Several QKD systems in the Rectorado node of the MadQCI. The Toshiba at the bottom is the actual Toshiba deployed for this quantum network scenario. QKD operation is supported by other quantum network systems such as optical and Ethernet equipment and IT systems. (b) A KEEQuant CV-QKD system on top and a ThinkQuantum DV-QKD below being tested at the laboratory. Several types of QKD exist, with different performance and network integration features, so an assessment is needed to choose the most suitable quantum system for each application. (c) A ZeroThird entanglement-based QKD system is commissioned with a vacuum pump; the UPS system on the right prevents vacuum losses in the event of a power failure. Its integration was more challenging, but the distribution of quantum entanglement will enable the “quantum internet”.
Source
The MadQCI Cloud Scenario: Quantum as a Service
Jaime S. Buruaga , Alberto Sebastián-Lombraña, Ruben B. Méndez, Rafael J. Vicente, Juan P. Brito, Laura Ortiz and Vicente Martin.
Entropy 2026, 28(3), 283; https://doi.org/10.3390/e28030283