Holding back could be a costly mistake – Why it’s time to take the Quantum-Inspired Optimization plunge

Main visual : Holding back could be a costly mistake – Why it’s time to take the Quantum-Inspired Optimization plunge

Technological progress is not linear: there are breakthroughs followed by periods of consolidation. Among many examples, the invention of the integrated circuit in the late 1950s led to all electronic and most mechanical devices becoming smaller and smarter in the following decades.

We are living in a similar period today, triggered by theoretical workaround quantum computing. When it can be made to work properly, quantum computing will make amazing things possible.

How about new plasma containment designs that break the logjam are currently holding back viable nuclear fusion reactors? Or personalized medicine that decodes not only your DNA (we can do that already) but also your unique microbiome (which is far beyond our current reach)? It’s also probable that genuine AI - computers actually thinking for themselves – will require quantum computing to become a reality.

The ability to optimize

One of the proposed applications for quantum computing is something called ‘combinatorial optimization’. This looks at exceptionally large data sets and seeks the ideal combination of factors to produce whatever solution you are seeking. A good example is drug discovery. Pharmaceutical companies are now using a technique called ‘virtual screening’ to assess whether a particular molecule could have the desired therapeutic effect. Rather than testing billions of candidate molecules in ‘wet chemistry’, they propose using quantum computing to do the heavy lifting in advance, then test the handful of really promising candidates that emerge from that.

Although enterprise-class quantum computing is still some years away, work on quantum algorithms now permits us to deal with certain classes of combinatorial optimization challenges – including virtual screening for drug discovery, using current digital computers, and without waiting for fully-operating quantum computers. This is Quantum-Inspired computing.

Businesses are ready to make this jump. Recent research by industry analyst Forrester, commissioned by Fujitsu shows that more than 8 in 10 business leaders are already familiar with combinatorial optimization challenges and are attempting to address them today. It also shows just under 9 out of 10 are already familiar with Quantum-Inspired computing as a solution.

Fujitsu is in the lead here with its Digital Annealer platform and the Quantum-Inspired Optimization Services based on it. With our partner Polarisqb, virtual screening for drug discovery is already underway, testing over one billion candidate molecules for a cure for Dengue Fever. This approach is not just more all-encompassing and more accurate, it’s also faster – reducing the time to reach pre-clinical testing from up to 48 months down to just eight.

In manufacturing, we are working with a major global automotive OEM to increase its PVC sealing capacity – and hence the ability to produce more cars with existing facilities, without the major capital investment into a new plant – by optimizing the path of sealing robots between and around vehicles. Current prototype quantum computing solutions addressing this challenge can optimize routes for about seven seams only. Working with Fujitsu’s Quantum-Inspired Optimization, the auto manufacturer is now calculating 64 seams per trip, increasing the choice of possible trip combinations to more than 10100, a number so big that it is far beyond the number of atoms thought likely to exist in the whole universe.  

And in financial services, we are helping organizations like NatWest bank in the UK and Main Incubator (part of Germany’s Commerzbank) with use cases such as financial portfolio optimization and stress testing, to provide both reduced risk and higher returns.

The upper limits of optimization

Until now, there have been strict limits on the complexity of what could be optimized. Some combinations of options were simply beyond the capability of current computing to process quickly enough to be worth the effort and cost, resulting in only partial optimization on simplified models.

Where the potential pay-off justified investigation with more complex models, the amount of computing time needed to run all the numbers might escalate to days, weeks, or even years for the most intricate cases. Again, the result has been a compromise, either by simplifying the model to save time or producing a full optimization that is quickly out of date because of the time needed to arrive at the answer.

These are the upper limits of optimization which have held us back until now.

Fujitsu’s Quantum-Inspired Optimization now makes full optimization possible in seconds – so fast that you can repeat the calculation as often as necessary and become not just optimized but ‘always optimal’.

Applying Quantum-Inspired Optimization to critical business challenges

Just what that means seems to be fully appreciated by business leaders, 85% of whom agree it could transform their business, according to the Forrester research, across an enormous range of potential applications.

When it comes to specific critical business challenges, there are clear opportunities from taking a radically new approach to product development, both in terms of previously unimagined modifications and enhancements, as well as providing the spark of innovation to develop new products, accelerate drug discovery and optimize investment in R&D.

Creating real-time solutions to complex problems using data and computing is also of great interest. For too many business processes, from financial portfolio optimization to job shop scheduling on a production line, today’s firms use point-in-time analyses because analyzing the data is a monumental challenge.

Other strong areas of interest are around optimizing physical-world problems, such as optimizing traffic flows for mobility platforms, first responders, and to relieve city congestion; solving logistics and warehousing problems like optimal route management, and optimizing the operation of manufacturing machinery.

No need to wait for quantum computing

Based on these research insights, it looks as if a coiled spring is about to be released by the availability of a proven, workable solution that does not require a five-year minimum wait for true quantum computing.

If the early adopters are already in the driving seat to become always optimal, what’s holding the rest back? More than seven out of 10 in the Forrester survey believe they lack the available skills. But that is now resolved with access to Quantum-Inspired Optimization Services– since the ‘S’ stands for Services. Getting started is quick and of negligible risk. As-a-service, cloud-based options do not require major IT investment or lengthy recruitment and training phases. Fujitsu already has all the algorithm expertise you need and means you prove a business value case in a matter of days.

Knowing exactly where to start is not the issue. This is technology so disruptive, no one yet has a clear picture of where it’ll end. Starting the thinking process is a critical need and we are keen to open dialog with you in a spirit of co-creation. The opportunities for you, or your competitors, are potentially transformational. Holding back could be a costly mistake.