Mr Duluc, quantum computers are seen as the new supercomputers that will revolutionize our economy. Do you share this view?

Philippe Duluc: Definitely, because quantum computers have a crucial capability. They execute special algorithms exponentially faster within milliseconds or seconds. Previous supercomputers take much longer for such variant-rich tasks or fail completely. In principle, these calculations can be solved in a simple way, but the computing effort skyrockets as soon as a certain number of variables is exceeded.

What application scenarios are there?

Philippe Duluc : You can find such tasks everywhere, which is why the huge potential of quantum technology can only be guessed at today. The tests already underway with quantum computers, for example by Volkswagen for traffic jam prediction in Beijing, illustrate the practicality of the technology. In this specific case, the movement data of 10,000 cabs is used to optimize their routes. The optimization of supply networks in electricity and water supply, the simulation of the folding process of proteins, image analysis and object recognition as well as optimization processes in IoT environments and machine learning show similar vast numbers of combinations and solutions.

... and in the industrial and automation sector?

Philippe Duluc : Quantum computers can make the most of their ability to process far more data in parallel than conventional computers, particularly in an Industry 4.0 environment. Think of predictive maintenance for control and monitoring systems in power plants and manufacturing environments, for example. Information from sensors on pumps and machines can be used to inform the maintenance service before a component fails. A quantum computer can process such volumes of data faster and more efficiently.

"Those who use the supercomputer revolution for their own purposes as early as possible will benefit from the coming changes."

© Atos

Can you briefly explain how a quantum computer works?

Philippe Duluc: Quantum computers work with quantum bits, or qubits for short. These have several states: 'zero', 'one' and any number in between - this is known as superposition. The second quantum effect is the entanglement of the qubits in a register, which causes their states to overlap. As a result, a change of state of one qubit is transferred to all entangled qubits immediately and independently of location - and can be measured. As a result of the calculation, these states can be translated into readable values of '1' and '0'. The qubit register calculates numerous possibilities in parallel in the superposition, which is ultimately the basis of the performance advantage of quantum computers. Atoms, electrons or photons are suitable as carriers for the qubits. Electrical circuits integrated into superconducting microchips can be used to store the qubits. However, the cooling required for the superconducting effect and the equipment for vibration protection make the technology very expensive. Companies can expect to pay tens of millions for a quantum computer.

Who is going to spend such sums on a technology that is not even ready for series production?

Philippe Duluc: No question, the price severely restricts the circle of users. All the protagonists from Google, IBM, D-Wave Systems, Intel and Microsoft, who are driving progress in quantum technology, are building their own prototypes. For example, Google only announced at the beginning of March that it was working on a processor with a computing power of 72 qubits - the current record. Nobody can say exactly when such powerful processors will be available on a mass scale. Serious estimates see quantum computers reaching market maturity in five to ten years at the earliest. However, thanks to cloud platforms from Google and IBM, there is already a chance that the user base for quantum computing will expand.

The second low-cost alternative is quantum simulators, which we also serve with the Atos QLM, the Quantum Learning Machine. It combines an ultra-compact computer with a universal programming language. This enables researchers and technicians to develop and test the quantum applications and algorithms of the future today. We offer systems with 30 to 40 qubits that can access 1 to 24 terabytes of RAM for quantum simulations. In the future, simulation platforms for different user groups will become increasingly important.

For whom and why are such simulation platforms interesting and important?
A simulation platform already gives engineers, software specialists, researchers and students the opportunity to familiarize themselves with the technology of quantum computers and, above all, to develop applications.
After all, it's not just about improving the hardware, but also about creating algorithms, programming techniques and development tools for quantum computers. Simulation platforms thus serve as the basis for applications in the fields of big data, artificial intelligence and IT security. The last point in particular is essential. This is because progress in the field of quantum computers also increases the security risk they pose. They will be able to crack asymmetric encryption techniques in particular, which work with public and private keys. One of the encryption methods under threat is the widely used RSA algorithm.

Where is the RSA algorithm used - and what is its security based on?
The RSA method encrypts online orders or EC card payments, for example. The security of the keys is based on the multiplication of large prime numbers. However, one of the strengths of quantum computers is their ability to break down prime number products into their factors. Even if these methods will continue to withstand quantum attacks for some time to come, it is necessary to develop new, quantum-safe security standards today. It can be assumed that cyber criminals, state hackers, secret services and other groups will gain access to quantum systems, preferably via the cloud. This poses a serious threat to entire industries and countries. However, the development of new security standards, for example to secure data transport between IoT endpoints and gateways in an IoT system, takes time.

What do you recommend that companies in the industrial and automation sector do in the near future?

Philippe Duluc : The security aspect in particular should also encourage industrial and automation companies to use quantum simulators. The initial focus would be on testing the strength of encryption techniques. The next step would be to develop solutions that offer protection against decryption attempts. Such algorithms are essential to protect customer and business data from unauthorized access. The second, equally important motive for companies to engage with quantum simulators should also be seen as a duty: Programming use cases in good time increases their own competitiveness.