One young man fascinated by the quantum field and one who pursued tennis and comedy are co-creating at Hitachi
Lee: Since I was a kid I enjoyed studying science and so I had a vague notion that I wanted to work in physics. When I was in junior high school I read books for general audiences (e.g. Blue Backs books published by Kodansha) which introduced me to the strange world of the quantum and showed me that there were interesting things out there. So that was what brought me, on entering university, to get involved in research into quantum computers that use light. I researched quantum computers in my master’s and my doctoral studies and earned a PhD. I was conducting research in the field of quantum entanglement, which won the Nobel Prize in physics in 2022, so I was really happy when the Nobel Prize was awarded right at the heart of my own student research discipline. Actually, once I finished my postgraduate studies I had the sense that I had done everything in research that I could, so when I finished my doctorate and was looking for a job, I wasn’t really hanging out for the quantum field. It was my sense that the research process I had been through would be useful whether it was the quantum field or not.
Utsugi: In elementary and junior high schools I generally liked studying, but I liked tennis more. Then when I got to senior high school I started to get really serious about tennis, but I gradually realized I didn’t really have much talent. I gave up on it as no good and changed to studying. Once I got into university I became engrossed in comedy, which I had always liked. I intended to make a living as a comedian, so I swapped to a comedy school. I actually went to the university administration office to tell them I was withdrawing and going to become a comedian, but I was exhorted not to withdraw and to just take a leave of absence, so I put my studies on hold and spent about two years on comedy. I realized I had no more talent at comedy than I did at tennis. Day after day I would get up on stage and no-one would laugh. I was broken hearted. Fortunately, I hadn’t withdrawn from university, so I was able to go back and study. Maybe I had more of a sense for study than I did for either tennis or comedy.
In the research room at university once I returned, I researched holographic displays. My motivation was really simple; it was fun to see things float in empty space. I finished a Masters and when I started to look for a job I heard that Hitachi was doing research into holographic memory, so thinking I could make use of my university research, I joined Hitachi. At the same time, there had been lectures at university about quantum computers and I had thought at the time they seemed interesting. I remember being told in lectures that there were a lot of Nobel prizes being given to the quantum research field and that made it particularly attractive.
From a different research field to quantum computers
Lee: When I joined Hitachi I took the stance that I didn’t mind what research field I was in. For five years after joining I belonged to a rare-earth magnet research team. At the interview before joining the company I had said I would do anything and I also felt that I could apply my fundamental research strengths to research in magnets, so I was able to go ahead with the research without any reservations. I was surprised that compared to university, the specializations and the ages of the researchers were really diverse. Particularly at Hitachi, where there I was as a physics specialist sitting next to someone who was specializing in bioscience, I really got a sense of the diversity of research. Even on the magnet team, team members had all sorts of specializations, and the way the leaders melded their teams together and really vitalized the research activity was something I didn’t see at my university.
Utsugi: The area I got involved in when I joined the company was research into holographic memory. In 2013 when I joined, it was just when the Blu-ray optical disk had come out and Hitachi was concentrating on holographic memory as a candidate for creating the next generation optical disks capable of recording much more data. I was with the project for about three years, during which time I was able to achieve some measure of research success, including receiving an award for the demonstration of a prototype machine in the NHK Science and Technology Research Laboratories, STRL Open House. I continued a bit longer with technological development of holographic memory and optical inspection devices, but I wanted to get involved in future technologies and alongside my job I began to study quantum computers for myself. I was undertaking self-motivated research into quantum computers, on my days off convening a quantum information study group with people from outside the company and writing a series of articles for a web media magazine, CodeZine (published by Shoeisha), for software developers. In 2019 Shoeisha published my book E de mite wakaru ryoshi konpyuta no shikumi (A visual guide to how quantum computers work). I don’t really have any particular hobbies, so concentrating on the study group and writing on weekends, I was able to publish the book about five years after beginning my studies.
Lee: I researched permanent magnets for about five years and once we had succeeded in creating a magnet with less rare-earth, the team was ultimately disbanded and absorbed. For about a year during that time I had been applying the knowledge from my university days to repeated proposals for quantum research. In my in-house search during that time for people interested in quantum computers, I started to encounter Utsugi san. While I was repeatedly submitting project proposals, patents and papers and other outcomes would come to light. As a result, after more than 20 years since the establishment of the Hitachi Cambridge Laboratory in England, which had been researching quantum computers, a team to develop a large-scale quantum computer was also set up in the CER in Japan. After that, I came to be working with Utsugi san on developing a quantum computer.
Utsugi: There are times when, like Lee san, you put up a proposal because you want to work in the quantum field and it gets accepted, but there are also times when that doesn’t happen. I had also been putting forward a proposal for a research topic related to the quantum field, but I couldn’t get it accepted. Once I understood that it wasn’t enough to just put in a proposal, I kept in touch with Lee san and continued to negotiate, working into the proposal the notion of the need to develop quantum technology for the society of the future, and eventually we got to where we are now.
Two research projects to create silicon quantum dots in a matrix structure
Lee: Our R&D is based on the idea of using silicon quantum dots as a way to achieve a quantum gate computer. Quantum computers calculate using minimum-sized units of quantum information known as qubits. Computing proceeds by gathering lots of qubits, controlling them and causing them to interact. There are a range of methods that have been proposed to achieve practical implementation of a quantum computer, including super conduction, ion traps, and optical implementations, but it’s our view that the silicon qubit method we are pursuing has advantages on multiple fronts.
Hitachi’s silicon quantum dot initiative sets up a type of bucket called a quantum dot, where only one electron is introduced into a semiconductor. Electrons have spin, which is a tendency to take one of two values, up or down, and the technology treats the direction of spin of the one electron the quantum dot contains as information and makes it a qubit.
The benefit of a silicon quantum dot is that the same semiconductormanufacturing technology as is used for CPUs and memory can be applied, so that it’s easy to replicate and integrate. An additional feature is that the peripheral circuit controlling a silicon quantum dot can be built on the same silicon substrate. The issue in a quantum computer is how to move a lot of qubits, but we believe that the silicon quantum dot will be effective in the future from the perspective of scaling-up. At this stage, however, we’re working on increasing the controllability of single qubits.
There are two aspects to our research. One is to integrate and move silicon quantum dots two-dimensionally, and by making a matrix structure, to develop a device that makes it possible to select and control electron-containing quantum dots.
Utsugi: To explain it simply, conventional methods require one connection for one qubit. So if you have 100 qubits, you need 100 connections and 100 controllers. Conversely, with Lee san’s matrix-style device, which can be controlled by a peripheral circuit, it’s possible to control lots of qubits with a single controller.
Once the box to contain electrons, the quantum dot, has been developed, we will need to come up with a way to put one electron at a time into it. Several theoretical methods are being considered, but variation in device manufacturing actually has an effect and once things are scaled up, it is possible they won’t work. We have aimed to develop a way to be able to appropriately put an electron into a quantum dot, even at large scale. By combining these control enhancing technologies we think we will achieve the basics necessary to cause qubits to move in an integrated structure.
Preparing a “box” for 128 electrons on one chip
Lee: The device we have developed has quantum dots in a small area at its center. When lots of qubits are brought together to work as a computer, there will need to be a way to choose and control the electron in a particular quantum dot. It won’t be enough to just create lots of quantum dots, there will need to be capacity to select and control by at the same time creating a peripheral circuit.
We’ve succeeded in making on one chip, using the same process, a quantum dot array that has a two-dimensional matrix structure and a peripheral circuit and encircling electrodes. We have developed all of the dots, the design of the peripheral circuit and the process ourselves here at Hitachi, and have built it in a clean room at the R&D Group’s Kokubunji site. Being able to make quantum dots and peripheral circuits on the one chip will be extremely useful when silicon quanta are scaled up. The device’s quantum dot array comprises eight columns and 16 rows, which means 128 quantum dots can be achieved on one chip.
Two-dimensional silicon quantum dot array
Quantum chip containing in a single chip, a 16 x 8 quantum dot array and peripheral circuit
We have also succeeded in using the manufactured device to measure the characteristics of quantum dots. Variation occurs in matrix-structured quantum dots, with electrons not going in, only one going in, or two or more going in, depending on the voltage impressed. Only one electron needs to go in to make a qubit. By measuring the conditions of voltage applied to an electrode and how electrons went in, we found the conditions necessary for one electron to go in. By also using the peripheral circuit for control, we proved that from among 128 quantum dots we are able to measure a chosen dot.
Having said that, all we have achieved at this stage is the box and we have not got as far as discovering the conditions for simultaneously introducing one electron each to 128 quantum dots. Notwithstanding, if we are able to control the spin of each introduced electron in 128 quantum dots, the possibility of achieving a 128 qubit quantum computer will be in sight.
Conceptual change—applying electron conveyance technology to quantum dots
Utsugi: Lee san’s research has produced a box that configures quantum dots. Next we need to think of a way of introducing one electron at a time. We tried a different way than in the past of doing that.
What we tried was the application of a different technology, the single-electron pump. It involved making a pump that extracts only one electron from a location where there are many and as our reference we used a method in development since around 2007. It is technology that was originally developed from the perspective that if it were possible to control electrons and transmit them one by one at high speed, it would be possible to accurately determine a standard value for current. We came up with the notion of applying the technology to quantum computers and introducing only one electron to an electron box.
Diagram of an electron loading method applying a single-electron pump
The basic principle is as follows. You introduce one electron at a time to a quantum dot box by selecting one electron from a place where there are many and putting it into the dot. Then, if you send one electron at a time in a bucket relay to neighboring dots, you can put one electron at a time into all the dots.
We carried out an experiment to confirm the principle. It’s quite difficult to search for the conditions for sending just one electron, so by searching and gathering lots of different data we finally succeeded in discovering the conditions for sending a single electron using a single-electron pump. We proved by experiment that only one electron flows in our device.
Then, we sent single electrons to adjacent quantum dots in a bucket relay. To best move the electrons there is a need to control the voltage between dots with a type of shutter to shield and transmit electron movement. By controlling the voltage with several gates we were able to confirm the motions of pump and shutter.
We’re at the stage where we have been able to house a single electron in a dot using a single pump and we’ve also been able to make a conveyance shutter in a bucket relay to adjacent dots, but we’re yet to achieve spin control. Nevertheless, the basics of a method for installing one electron at a time into a large-scale quantum dot array are almost complete. If we can install electrons in an array by bucket relay and then achieve control of the spin, we will have come closer to practical implementation of a large-scale quantum computer.
In this research we collaborated in our experiments with Tokyo Institute of Technology (Tokyo Tech) and were able to release the experiment results very quickly. I remember how hard it was to get things up and running because I started with the unfamiliar, including handling of experimental devices—I was in a different experimental environment than when I had worked in optical technology, and it was my first time experimentally verifying quantum computer related technology.
Lee: In creating a one chip array, when electron movement doesn’t go as ideally it should, it’s my impression that it’s difficult, but interesting to achieve a win. Trying to simultaneously make quantum dots and the peripheral circuit in the same process is a world-first, so we weren’t sure it would go as we intended, but I think we’ve overcome the difficulties and results are in sight.
The value of researching at Hitachi with its wealth of human talent
Lee: In the sense of research and development into quantum computers, both Utsugi san and I are responsible for the bottom layer; the device. To achieve practical implementation of a quantum computer requires very many layers of technology to be integrated, including software and architecture. It’s clear that can’t be done by one person.
However, in Hitachi, there are not only researchers like us who know about hardware, there are many outstanding researchers from diverse backgrounds. We have got to where we are because those people have helped us. Hitachi is unique because it has a galaxy of talent.
On top of that, to create a high quality quantum dot we sought to learn from the professors at Tokyo Tech. I think that being able to draw on the strengths of all those people is how we’ve been able to come up with this original technology, where we put quantum dots into an array to make the quantum dots into a quantum computer and open a path to scale-up.
Utsugi: I think another feature of the Hitachi R&D Group is that the Group includes people who are passionate about technology. They aren’t the types who just do their jobs. If it’s quantum computers, then they have the enthusiasm for and the conviction to absolutely want to make that quantum computer.
Lee: I think what’s attractive is the ability to do development in a team where there is a variety of human talent. Also, if you can tell the story of how absolutely amazing something will be if it goes well, the environment is such that you can make a few mistakes and you’ll still be forgiven (laughs). That’s something to be really grateful for when you’re in research and development.
Lee: I think also that there’s a perception that there are a lot of serious people among Hitachi researchers. Those working on quantum computers too get seriously involved in challenging topics. Researchers with a wide range of knowledge apply their individual abilities in all seriousness and one hundred per cent to research and development, and I think that’s the strength of the Hitachi R&D Group.
LEE Noriyuki
Senior Researcher
Center for Exploratory Research
RESEARCH & DEVELOPMENT GROUP
Moved by Soseki’s conclusion of a century ago, that you should do the things you want to do
I have read the collection of Natsume Soseki’s lectures, “Watashi no kojin-shugi” (My Individualism and the Philosophical Foundations of Literature) (Author: Natsume Soseki, published in Japanese by Kodansha Academic Library) multiple times. “Gendai Nihon no kaika” (The Civilization of Modern-Day Japan) was in our Japanese textbook, and that caused me to read the book because I was interested in his other lectures. I like Soseki’s novels and had begun reading them when I was a kid.
In that book Soseki explains that we should do the things we want to do, rather than what others tell us to do, while respecting what others want to do. Such a statement presents no surprises nowadays, but the fact that he arrived at that conclusion by himself a hundred years ago left a profound impression on me and I’ve read it over and over again. Researchers are the same—even if we have the same objective, depending on the person, the approach and the methodology will likely be different. Even so, persevering with your own way of doing things while respecting the way others go about it is an important mindset, I feel.
UTSUGI Takeru
Senior Researcher
Center for Exploratory Research
RESEARCH & DEVELOPMENT GROUP
I recommend Shinsoban—Keisankiya kakutatakaeri (Special edition—How calculator makers battled) (Author: Satoshi Endo, ASCII Media Works)
In a single volume the book recounts the history of computer development in Japan. The stories are enthusiastically told by 26 pioneers who themselves competed in the world-wide competition to develop computers. Among the episodes, the one who particularly made a profound impression on me was Toshio Ikeda, who founded Fuijtsu’s computer business. Ikeda appears to have been the type of person for whom there was no precedent; he often didn’t come to work, for example. Many people nevertheless recognized him as a genius and followed him, so that ultimately he rose to the top of the computer business and led the business. It’s a book that really gives courage to people who build computers.
