Skip to main content

HumanQ for quantum computing: about us

Italian quantum emulation company within the Polo Innovativo ecosystem, thanks to EQM technology from MCG


This page introduces HumanQ, a company founded at the beginning of 2026 that offers quantum emulation, assessment of business readiness, training and pilot projects. It describes the EQM platform, its scaling limits, the relationship with customer data and the collaborations under way.


Who is HumanQ?

HumanQ is an Italian quantum computing company, founded at the beginning of 2026 within the Polo Innovativo ecosystem, that brings quantum emulation to researchers, businesses and institutions. It uses the EQM platform, owned by MCG, founder of the Centro Italiano di Emulazione Quantistica, and is the exclusive worldwide seller of EQM technologies, so that quantum algorithms can be tried without dedicated hardware.

The company is based in Brescia, at Polo Innovativo, which supports it both as incubator and as shareholder. The director and legal representative is Valentina Cordeschi, and the lean technical team today brings researchers and physicists together with software architects.

The daily work revolves around a simple idea, making quantum computing usable right away for those in research, business or public administration, starting from what classical machines can already emulate with accuracy.


What does HumanQ do for quantum computing?

HumanQ works on five fronts that support one another. Quantum emulation lets you run quantum circuits and algorithms on classical machines, and it is used to design and verify them today, without waiting for suitable quantum hardware. The quantum readiness assessment measures how prepared an organisation is and produces a roadmap, and it is the step through which a company finds out which skills and which databases to start from. Training and outreach teach the basic concepts, from superposition to entanglement, to the people who will have to use them.

Industrial pilot projects bring quantum computing to a real, bounded and measurable use case, so that the value can be seen before any larger investment. The development of quantum algorithms finally answers specific customer problems, for example optimisation or simulation. In every case HumanQ develops quantum emulation projects, to keep the focus on what is achievable today.


What is the EQM platform and who owns it?

EQM stands for Modular Quantum Emulation (Emulazione Quantistica Modulare) and is a platform developed by MCG that makes it possible to design, run and analyse quantum circuits through high-performance software emulation, drawing on classical computing architectures. The modular structure separates the logic of the problem from the implementation of the underlying quantum gates, which makes it possible to change the computing tools without rewriting the algorithms.

The EQM platform used by HumanQ was unveiled as a world premiere at the Centro Italiano di Emulazione Quantistica on 7 July 2026 at the Auditorium Acquario in Genoa. HumanQ is the exclusive worldwide seller of EQM technologies and offers them as machines that customers install on their own premises, as explained in the section on data sovereignty.


How many qubits can be emulated with EQM?

The number of qubits a machine can emulate depends on the GPU and on the memory associated with each GPU, which is why the EQM family comprises four machines, from EQM1 to EQM4. With the state vector technique you can reach up to 34 qubits on workstations and servers. To reach 40 qubits with the same technique you need a cabinet of several GPUs linked together, that is data centre infrastructure. With tensor networks you scale up to 100 qubits, again in a data centre, and the probabilistic accuracy of the computation falls to 87% against 99.9% for the state vector.

Emulation capacity of EQM by computing technique
TechniqueQubitsInfrastructureProbabilistic accuracy
State vectorup to 34workstations and servers99.9%
State vectorup to 40cabinet of several linked GPUs (data centre)99.9%
Tensor networkup to 100data centre87%
Proportional bars: 34 qubits on workstations, 40 in a data centre with state vector, 100 with tensor network
Scale of the qubits that can be emulated with the EQM platform and the related probabilistic accuracy of the computation.

What is the difference between state vector and tensor network?

In the state vector method the computer keeps an amplitude for every possible configuration of the qubits, which is why the result is faithful to the mathematical model, with a probabilistic accuracy of 99.9%. The price of this fidelity is memory, because every added qubit doubles the space required.

In depth. The state of n qubits requires 2n complex amplitudes. In single precision, with eight bytes per amplitude, 34 qubits occupy over 130 gigabytes and 40 qubits over 8 terabytes, a difference that explains the move from the workstation to the data centre.

Tensor networks follow a different route. They represent the state as a network of smaller tensors linked to one another, and in this way they compress the correlations between qubits. The compression works well when the entanglement of the circuit stays limited and it introduces an approximation, which is why with this technique EQM reaches up to 100 qubits with a probabilistic accuracy of 87%. The value depends on the type of circuit and should be read as a declared trade-off between scale and accuracy.


What is the difference between quantum emulation and simulation?

An emulator replicates both the hardware and the software of a system, so that it works exactly as the mathematical model predicts, whereas a simulator only creates an abstract model or a virtual environment that imitates its behaviour. The distinction matters in quantum computing because emulation makes it possible to verify an algorithm as if it were running on a quantum machine, while simulation offers a simpler representation of it.

HumanQ develops quantum emulation projects and goes deeper into the comparison on the pages dedicated to quantum emulation and quantum simulation.


Why is the error of a logical qubit almost zero in emulation?

The logical qubit is the ideal qubit the algorithm reasons about, and it differs from the physical qubit that implements it in a real device and is subject to errors. In emulation the error of a logical qubit is close to zero, because its state is computed in memory on GPUs and does not live in a component exposed to disturbances.

Decoherence and quantum noise are phenomena typical of physical machines and in emulation they do not occur spontaneously, so they are recreated deliberately when it is necessary to study their effect on an algorithm, as the page on quantum error correction explains. To explore the concept of the quantum bit further, you can start from the page on the qubit.


What does data sovereignty mean?

Many research, industrial or institutional datasets often cannot leave the perimeter of those who produce them, for reasons of confidentiality, regulation or strategy. For this reason EQM can be installed as a stand-alone machine at the premises of researchers and authorities, and the computation takes place entirely within their perimeter.

The main advantage of this model is data sovereignty, that is the ability to keep control over where data resides and who processes it. The technology used does not come from outside the EU.

Diagram: data, stand-alone EQM machine and results stay within the customer perimeter, with no link to the cloud
With a stand-alone machine the computation takes place within the perimeter of the data owner.

Where do you start with quantum computing?

Anyone who already knows entanglement, superposition and interference can start working in emulation within a few weeks. Anyone who does not master these concepts and does not develop in Python has a longer path ahead, of at least six months, during which the basic culture has to be built.

The pages on quantum superposition and quantum entanglement offer a first entry point, while the quantum readiness assessment helps decide which skills to develop first. In business practice the path rests on four levers, namely quantum literacy, database development, emulation and development of quantum algorithms.


Who does HumanQ work with?

HumanQ operates with an Open Innovation model that involves universities, businesses, startups and institutions. It collaborates formally with Professor Marco Armoni, father of the EQM technology and a leading figure in quantum computing as well as author of numerous publications on the quantum computer.

With Wise Solutions it works on a plan for a future heuristic scheduler, while with SpaceTaly it carries out research on textile materials for the space economy. These activities show how emulation applies to concrete industrial and research problems, and the pages on quantum machine learning and quantum computing in finance show many other fields of use.


What are the next steps for HumanQ?

HumanQ is taking part in the public consultation of MIMIT, the Italian Ministry of Enterprises and Made in Italy, on quantum technologies 2026, with the aim of qualifying as a national supplier of quantum services, and in its answers it describes precisely the capabilities actually available. The company also intends to strengthen its activity in quantum sensing and metrology.

For industrial problems, the page on the applications of quantum computing collects the sectors of interest, from health to finance.


How can you contact HumanQ?

For a demo of the EQM platform, a quantum readiness assessment or a pilot project you can write to info@humanq.it or use the form on the Contact page. HumanQ is based at Polo Innovativo, at Via delle Scuole 1/G, Brescia (25128), Italy.

Request an EQM demo


Are you a researcher?
Experience Quantum Computing firsthand

Universities, research centers, companies: contact us for a platform demo or to explore collaboration on dedicated projects.

I declare that I have read the information notice regarding the processing of personal data
CAPTCHA
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.