Fujitsu, 1QBit work on quantum-inspired AI cloud service

Fujitsu and 1QB Information Technologies are collaborating on applying quantum-inspired technology to the field of artificial intelligence (AI), focusing on the areas of combinatorial optimization and machine learning.

The two companies will work together in both the Japanese and global markets to develop applications which address industry problems using AI developed for use with quantum computers.

This collaboration will enable software developed by 1QBit for quantum computers to run on a "digital annealer," jointly developed by Fujitsu Laboratories and the University of Toronto. A digital annealer is a computing architecture that can rapidly solve combinatorial optimization problems using existing semiconductor technology.

Over the last four years, 1QBit has developed new methods for machine learning, sampling, and optimization based on reformulating problems to meet the unique requirements of interfacing with quantum computers. The combination of Fujitsu's computer architecture and hardware technology, and 1QBit's software technology, will enable advances in machine learning to solve complicated, large-scale optimization problems.

Fujitsu has systematized the technology and its experience with AI under the name of Zinrai, which has developed over the course of more than thirty years. The platform will support customers in using AI and will be available as the Fujitsu Cloud Service K5 Zinrai Platform Service.

Within 2017, Fujitsu will offer the results of this collaboration as an option in the Fujitsu Cloud Service K5 Zinrai Platform Service Zinrai Deep Learning, a Zinrai cloud service.

In the future, the two companies will provide a variety of services that combine 1QBit's software and expertise in building applications which benefit from the capabilities of quantum computers, with Fujitsu's hardware technology, its customer base — the largest in Japan — and its versatile ICT capabilities, including AI.

The partnership aims to contribute to the creation of new businesses and the transformation of existing businesses by introducing new solutions to the computational challenges facing customers in a variety of fields, including finance, life sciences, energy, retail and distribution. According to Hirotaka Tamura, Fellow, Fujitsu, applications range from finance (e.g., portfolio optimization) to techniques that can be used for drug discovery (i.e., search for molecules).

The following comments were made by Tamura in response to Enterprise Innovation's queries on this initiative:

What sort of applications can be foreseen?

At Fujitsu we think that one of the important applications of our quantum-inspired hardware (digital annealer) is combinatorial optimization problems, which are often intractable. Our assessment, however, is that the current quantum annealer is too limited in terms of the size and problem-mapping accuracy to handle real-world problems. As a result, we have been developing a quantum-inspired digital annealer that has higher usability than that of commercially available quantum annealers. In contrast, 1QBit has been devising various ways to handle real-world problems with an existing quantum annealer, but has been suffering from the limited size of the annealer.

Now that 1QBit and Fujitsu are collaborating, Fujitsu has access to 1QBit’s technique to handle problems that potentially have markets, and 1QBit has access to the hardware that expands the size of the problem they can handle to a realistic level.

Also, researchers are now looking into the application of quantum and quantum-inspired hardware to machine learning and AI. AI is already giving great impact to society, and the combination of our digital annealer and 1QBit’s software can provide unique and crucial solutions to the area because it can speed up the learning in deep networks different from conventional deep neural networks.

Which verticals might this disrupt and in what ways?

Our observation is that there are combinatorial optimization problems behind many difficult computational tasks. Many hard combinatorial optimization problems and computation for learning in AI are such examples. Quantum-inspired hardware-software complexes will penetrate into markets starting from such areas. The unique nature of our approach, which is the effective use of parallelism through stochastic search backed by the latest knowledge of statistical physics, will bring the performance of our hardware to the level that ordinary processor solutions cannot reach. This would be disruptive for conventional processor solutions.

Also, the speed and the size of our near-future hardware will potentially change the way of performing jobs of, say, financial advisers or material-discovery scientists. As shown in these examples, the higher speed with larger size that cannot be obtained by conventional processors will create new applications and new markets.

What are some of the longstanding, intractable computer science problems it might solve?

We currently believe there are intractable problems no matter what hardware is applied. We, however, will be able to solve problems that have been thought intractable mainly due to the time constraint. If we can provide hardware that gives a ten-to-the-fifth-order speedup, for example, a task that takes one year will be finished in about 5 minutes. We are going to achieve this goal by using our approach that provides speed-up means not dependent on Moore’s law, which will be ending shortly. In other words, we are providing methods to sustain the performance growth of computing in the post-Moore era, which effort is certain to solve longstanding, hard issues in computer science.