Programming on quantum computers: what is it and how does it work?

Programming on quantum computers is emerging as a new frontier in the world of IT.  It exploits the principles of mechanics to process information in radically different ways compared to those used by conventional computers.

On the one hand, this new technique offers extraordinary potential, but on the other, it requires significant changes in our approach to software programming.

Basics of quantum computing

Unlike traditional computers (which use bits with a value of 0 or 1 as their basic unit of information), quantum computers use qubits.  Thanks to a phenomenon known as superpositioning, they can represent both state 0 and state 1 simultaneously; this special ability enables the system to process a large number of possibilities at the same time.

Furthermore, entanglement allows qubits located some distance from each other to be correlated so that the state of one instantly influences the state of the other, regardless of the distance between them.  These features mean that quantum computers are able to deal with complex problems with a level of speed and efficiency unmatched by conventional systems.

It is therefore apparent that quantum programming is quite different from traditional programming in the ways listed below:

• in traditional systems, algorithms are determinist and follow a linear sequence of instructions;
• quantum algorithms however, are probabilistic and exploit parallel processes on multiple states (thanks to superpositioning and entanglement). This requires a new approach to software development, as traditional Boolean logic is not sufficient for the successful management of qubit operations.

Examples of quantum programming languages

In order to develop quantum applications, specific programming languages have been created:

• Q#: developed by Microsoft, it is part of the Quantum Development Kit and allows programmers to simulate and execute quantum programs. When integrated with Azure Quantum, it offers tools to estimate the resources required and test algorithms on real quantum hardware;
• OpenQASM: used by IBM, it is a quantum computing assembly language which describes the circuits that can be used on IBM’s quantum processors;
• PyQuil: a Python library which facilitates the writing of quantum algorithms compatible with Rigetti Computing processors, simplifying the integration of quantum and conventional code;
• PennyLane: this library focuses on automatic quantum learning and provides tools to blend quantum algorithms with conventional machine learning techniques.

In addition to specific languages, there are several tools and frameworks which support quantum computer programming:

Qiskit: an open-source project developed by IBM, which enables quantum circuits to be created and manipulated using Python.  It offers simulators and access to real quantum devices via cloud;

Strawberry Fields: an open-source library developed by Xanadu Quantum Technologies to design, simulate and optimise optic circuits with continuous variables, with simulators based on different representations of quantum mechanics;

t|ket>: a programming environment and optimising compiler, developed by Cambridge Quantum Computing, which supports several backend hardwares and simulations.

Applications of programming on quantum computers

Programming on quantum computers is finding applications in a range of environments, with the potential to revolutionise entire sectors.  Here are some examples:

• simulation of molecular and chemical systems: one of the most promising environments for the use of quantum computers is computational chemistry. Thanks to their ability to manage huge quantities of quantum states at the same time, these systems can simulate the behaviour of complex molecules with great accuracy, thereby accelerating the discovery of new materials and pharmaceuticals;
• combinational optimisation: complex optimisation problems such as the management of logistical routes or efficient resource distribution can reap huge benefits from quantum programming. Certain algorithms, such as Grover’s, offer considerable advantages over conventional approaches;
• quantum finance: the financial sector can exploit quantum calculations to create more accurate predictive models and extremely advanced risk simulations;
• artificial intelligence and machine learning: automated quantum learning or Quantum Machine Learning, is an emerging area which could lead to significant improvements in the ability to process and analyse large quantities of data;
• cybersecurity and post-quantum encryption: traditional encryption based on RSA and other conventional algorithms could be vulnerable to quantum computers, which have the ability to decipher the encryption keys currently in use very quickly.

The main quantum algorithms and their current limits

In order to exploit the potential of quantum computing, several specialised algorithms have been developed.  Some of the most important include:

• Shor’s algorithm (factorisation of whole numbers): Shor’s algorithm enables the factorisation of whole numbers much more quickly than conventional algorithms. This is especially relevant to encryption, as many IT security systems are based on the dificulty of factorisation;
• Grover’s algorithm (non-ordered research): Grover’s algorithm enables the location of an element in a non-ordered database in quadratic time compared to a conventional algorithm;
• Variational Quantum Eigensolver (VQE): VQE is a hybrid quantum-conventional algorithm which helps to identify the minimum proper values of complex systems, necessary for the simulation of materials and molecules;
• Quantum Approximate Optimisation Algorithm (QAOA): QAOA is an algorithm designed to solve problems involving combinational optimisation using a hybrid quantum-conventional approach.

Although quantum computers offer enormous potential (as described above), there are still several challenges to be faced in order to make their largescale adoption possible.  The main ones are:

1. errors and quantum noise: current systems are subject to errors caused by decoherence and system noise, which reduce the reliability of the calculations;
2. difficulty of scalability: quantum computers with many stable qubits are still in a developmental phase;
3. elevated costs: quantum infrastructures are costly to develop and maintain, which limits their uptake;
4. limited algorithms: many quantum algorithms are still in an experimental phase and have not yet achieved largescale industrial application.

All in all, we can conclude that programming on quantum computers represents an authentic IT revolution, with the potential to solve problems that conventional computers are unable to handle within a reasonable timescale.

While there are still many technological challenges to be overcome, the rapid evolution of hardware and algorithms is making the dream of quantum computation accessible to everyone.  Knowing how to program these systems has therefore become an increasingly sought-after skill for the future of IT and computational engineering.

Translated by Joanne Beckwith