2 min

Accenture has been granted a patent related to quantum computing. The patent describes a method of using machine learning to determine which workloads or problems are most suitable for quantum computers, and which are more suitable for standard, digital computers.

According to Accenture, the patent is useful for early adopters of quantum computers because it helps them to better balance costs, writes Silicon Angle.

Quantum computing

Quantum computing is different and much more powerful than traditional computer architecture. This technique has the potential to solve extremely complex problems that are impossible to solve for today’s computers, or that can only be solved after years.

The main difference between the techniques is that in quantum computing, processing can take place in several states at the same time. Traditional computers use binary digits, also called bits, which are represented as 1 or 0. Quantum computers use qubits that can be represented as 1, 0 or both at the same time. Qubits can also use superdense coding, where they can hold two bits at the same time.

Another important difference is the ability of qubits to correlate with each other. Therefore they are all aware of the state of the other qubits. As a result, quantum computers grow exponentially in strength when qubits are added.

Challenges

Quantum computers are therefore much more powerful than today’s computers and can solve problems faster. A challenge for many early adopters is probably that it is not clear when to use quantum, and when to choose a combination with traditional computers, says Marc Carrel-Billiard, senior managing director at Accenture Labs.

The company’s new patented machine learning module should help early adopters to learn how to handle the new technology. The company calls this computational variety.

By determining when and how to use the power of quantum computing, the system can help to perform computational tasks in the most efficient and cost-effective way. In addition, the described module can learn to prioritize certain tasks. And if more advanced and efficient quantum and classical systems are introduced, the quantum computing machine learning module can adapt to this, says Accenture.

This news article was automatically translated from Dutch to give Techzine.eu a head start. All news articles after September 1, 2019 are written in native English and NOT translated. All our background stories are written in native English as well. For more information read our launch article.