Quantum Software

Quantum Algorithms, Software and Theory (QAST) encompasses research that is focused on the theoretical aspects of quantum information science. This may be in developing and analysing new or existing quantum algorithms, communications or sensing protocols, developing tools for quantum error correction or error mitigation, building compilers or performance analytics tools or developing new programming interfaces, languages or new ways to interact with quantum computers. QAST also includes research into how second generation quantum technology will impact our current information processing systems, such as how threats in cryptographic systems from quantum technology may necessitate new methods of encryption and even fundamental theoretical research regarding the nature of quantum computing, communications or sensing technology.  This includes what quantum  technology can do and how it relates to more foundational concepts in physics, mathematics, computer science and chemistry.

• Cybersecurity: The security of the internet is underpinned by public key cryptography. Public key cryptography is based on mathematical functions that are difficult to compute but easy to verify. Essential to understanding the threat posed by quantum computers towards cybersecurity is establishing the size of a quantum computer that is capable of attacking modern public-key cryptosystems. Currently, the best estimate for attacking RSA 2048 with a quantum computer requires a device that has approximately 20 million qubits. While this is a large number, this number has fallen by a factor of 1000 in 10 years due to advancements in algorithms, compiling, and error correction. If this can be further improved by an order of magnitude through theoretical developments, RSA 2048 would be under threat within a decade, assuming the projected manufacturing timelines of IBM. Due to the danger posed by “harvest now, decrypt later” attacks, there is an urgent need to deploy “post-quantum” security schemes that cannot be exploited by quantum computers. The US is pioneering the transition to quantum-safe cryptography via a presidential executive order and standardizing post-quantum cryptography protocols. Australia must follow the US lead and set the requirements to update security protocols for critical systems and data.
• Simulation: Quantum simulation is a set of algorithms for simulating the dynamics of quantum systems, which becomes classically intractable as the system size grows. Quantum simulation is also a foundation of quantum speedups for quantum chemistry, material science, and even quantum linear algebra. However, simulating practical problems that are unambiguously beyond the reach of classical computers requires quantum computers capable of full-scale error correction. Despite the required resources, it is likely that quantum simulation problems will be the first to demonstrate the utility of quantum computation for industry-relevant problems on a large scale, and there is some hope that some advantage over classical computers can be obtained without fault-tolerant devices.
• Optimization: Optimization algorithms are utilized in almost all computational problems. Even seemingly distinct problems in machine learning and AI or computational chemistry are in practice solved through optimization. From the point of view of computer science, most of these problems are NP-hard, i.e., not solvable in general by computers (quantum or classical) at scale. However, real-world scenarios tend to have additional structure that allows for practical solutions in many cases. Such applications are often developed through competitions such as the grid optimization competition. Given the potential impact of even mild improvements to optimization algorithms, there has been a focus on quantum algorithms for optimization since the 1990s. Current research indicates potential “quadratic” scaling improvements for quantum computers.
• Machine Learning and AI: Processing big data has been identified as a potential application for quantum computers, and there are data-processing tasks that can be performed more efficiently on quantum computers. However, the input and output of the data remains a bottleneck for quantum computers, erasing most known quantum speedups for machine learning and AI. While quantum advantage in terms of computational time is not expected in the foreseeable future, there are examples where a quantum computer needs exponentially less data for learning. This is most pronounced when learning from quantum data, such as quantum states produced by multi-qubit quantum sensors.
• Performance Analytics: Quantum benchmarking and performance analytics is a rapidly expanding field to both assess the utility of quantum computing for simulation, but also build new techniques to reduce the physical resources to realise these applications. Both the United States and Europe have initiated region-wide programs to build out the tools and capabilities to accelerate the ability of quantum computers to achieve advantage across all of these broad applications.
• Quantum Key Distribution: Quantum Key Distribution is often cited as the quantum solution to the vulnerabilities posed by quantum computers to cryptographic systems. By replacing computational hardness assumptions with a physical system based on quantum mechanics, we can guarantee by the laws of physics a method for distributing random but correlated bit-strings between two users that could be used to establish a secret key. Quantum Key Distribution remains the only commercially viable quantum protocol – in communications or computation – that is available today. However, the technological capability of QKD systems compared to ‘adequate’ non-quantum solutions that are also available has limited the up-take of commercial QKD outside of a few select government and private sector actors.
• Quantum Architecture Development: How do we design computers, communications and sensing systems that don’t actually exist?  This is the responsibility of those who work in quantum architecture development.