Finally Created by Collaboration with the University of Toronto
(Continued from Part I of the interview.)
Digital Annealer solves combinatorial optimization problems at high speed using a digital circuit inspired from quantum mechanics. To give shape to Tamura's ideas, Fujitsu and the University of Toronto created a unique mixed research and development team in 2016.
"We began our exploratory research in 2014, and we started considering possibilities to develop hardware. At the start of 2016, we asked Professor Ali Sheikholeslami of the University of Toronto in Canada, with whom we had previously collaborated in other research projects, to come to Japan to help us with our research.
"From the University of Toronto came Professor Ali and two students who were skilled designers. On the Japan side, there was one hardware designer, one analog circuit designer, and a team of three quantum cryptography experts. This was a unique mix of people with diverse talents.
The team worked well together and we continued our research, creating a prototype to present in a demonstration at a research strategy meeting held in 2016."
This demonstration was enthusiastically received, and since then, the team's research and development on Digital Annealer has been on a steady trajectory.
Meeting Professor Ali, an Indispensable Partner
In Professor Ali's research lab
"I first met Ali in 1998 while he was still a student at the University of Toronto conducting research for his doctorate. A company he had counted on to help him in obtaining measurement data and other tasks did not ultimately support him, so he used agents to attempt to find another company willing to collaborate, which is where Fujitsu Laboratories came in.
"I happened to be in the division that accepted his request, and we ended up lending Ali a hand at our research lab for about four weeks. What really impressed me about him was how even as a student, he had the ability to motivate our company's young engineers. He was very good at leading people and cooperating effectively to drive research forward. Besides being academically gifted, he can manage people and negotiate effectively.
"After a year or two, when I participated in an international conference, I ran into Ali, who was by then a professor at the University of Toronto. Next, we received another request about a research topic about a half-year later, to which I replied, 'I'll just include that in my current research topics' and forced it in (laughs). Since then, we have worked together on all kinds of research projects."
Research on the Digital Annealer: The Epitome of an "On-going" Process
When talking about the Digital Annealer, Tamura never neglects to mention that this is not the success story of a finished project, but rather a technology that is still undergoing research and development.
"The Digital Annealer's hardware implementation is just getting started―this technology is still on-going.
If you compare the current Digital Annealer to the history of the microprocessor, we are still at the stage in which we are 4-bit or 8-bit. Today's microprocessors can solve all kinds of problems like magic, but what is actually happening behind the scenes are simple calculations, such as addition, subtraction, and shift operations. The part that seems like magic is performed by extraordinary software. Similarly, to make the Digital Annealer appear to employ magic, we need technology to apply it--in other words, software. One effort we are making to render the Digital Annealer usable as soon as possible is our collaboration with the only company that is developing and selling quantum computing software, 1QBit."
Hardware That Can Provide the Optimal Method for Taking the Next Step
How would you like the Digital Annealer to be applied in the world as it becomes more widely used?
"There are many things in this world that require an optimal combination solution. This is where the Digital Annealer comes in―it can provide the optimal method for taking that next step. Decision-making is conducted based on some kind of value. For example, the value can be feelings, such as a sense of satisfaction, but ideally, the value is something that includes such feelings as hardware, which can determine what step to take next in the decision-making process. I think we must think about these usage scenarios as well."
Do Not Dismiss Optimism, a Sense of Humor, or New Ideas
When you speak about the innovative development of the Digital Annealer, you keep the conversation light in tone. Where does the driving force that motivates you to create something new come from?
"It's a little scary to think about how I might not come up with anything new even after thinking so hard. However, that is because I am trying to generate these ideas by myself; someone somewhere in the world may have already published something that draws on a similar idea. There may be someone out there in the business world who has thought of the idea but never really recognized it. If we combine the ideas that these people have, we might be able to solve these problems. Remember that even in the worst-case scenario, when you end up having to give up on your initial plan, that process is itself a part of the solution―one that is within reach.
"Out of a hundred ideas I might come up with, not a single one may succeed. If you get discouraged by that, then there is no way forward, so you must have some kind of optimism or a sense of humor, and consider the 99 failures you went through to be positive experiences.
In this research lab where an array of devices are scattered, new ideas are tested daily.
"What I personally consider to be important is to retain an atmosphere in which people are free to try something stupid. Although in actuality we seldom find ourselves in such ideal circumstances, the challenge lies in how much flexibility we can find in the process, and how much trial and error we can undertake. I do not think it is a good idea to have someone in a high position telling others to stop attempting unconventional ideas, thereby blocking new ideas from coming forth. In fact, it may be better for those in charge to say something silly now and then-- although there are of course limits to that kind of thing (laughs)."
Finally, we asked Tamura what challenges he would like to tackle while pursuing the innovative development of the Digital Annealer.
"The development of the Digital Annealer was a result of asking: what should the future of hardware be? In that sense, my hope is to discover something new that will help computers transition into the next phase, as was the case in the past with vacuum tubes, transistors, integrated circuits, and today's large-scale integration technology."
Since he was in elementary school, Tamura dreamt of becoming someone who creates new things.
That dream continues to this day.
Hirotaka Tamura, research fellow at Fujitsu Laboratories
Hirotaka Tamura graduated from the Department of Electrical and Electronic Engineering of the University of Tokyo in 1977. He completed his doctoral course at the same university in 1982, whereupon he joined Fujitsu Laboratories. After studying the Josephson device, the quantum effect device, and other topics, he began researching CMOS circuits.
He is a research fellow at Fujitsu, a Doctor of Engineering, a recipient of the 51st Okochi Memorial Prize, and an IEEE Fellow.