“Many pieces of the puzzle currently fit together”

Dr. Förtsch, what can quantum magnetic field sensors achieve at the human-machine or human-prosthesis interface?

It has given us, for the first time, an instrument with which the biosignals of the human body can be recorded touch-free and made available in the digital space. The sensors require no contact with the skin to detect muscular movements. What sounds abstract opens up very concrete possibilities: Who wouldn’t want to put their smartphone down and still be able to use it? Who wouldn’t want to use the PC without a mouse and keyboard? The eye shows where we want to move the mouse. It is controlled by muscles whose biosignals can be tracked non-invasively by our sensors, which could be embedded for that purpose in the glasses frame, for example. It would also be conceivable in the future to track and interpret hand gestures or arm movements. This is already being done in part with camera systems, but these are computing-intensive and consume so much power that today’s batteries soon reach their capacity limits. Our quantum magnetic field sensor does not need to process image data, but instead uses native signals from the human body. We therefore also talk about “native sensing”, with which we make these signals available in the digital space. This enables new, efficient approaches to controlling prostheses, exoskeletons, and cobots, as well as PCs, smartphones, and many other devices.

If your sensors measure picotesla magnetic fields—which corresponds to one millionth of the Earth’s magnetic field—are they not very susceptible to signal interference?

We have, of course, also asked ourselves this question and found technological answers. Before I explain them, let me start with one thing. It is our maxim that our products must firstly be free of any cooling and secondly be suitable for everyday use, in other words, usable outside of ideal conditions in the laboratory. Only then do we start with product development. As far as the magnetic field sensors and the risk of signal overlap due to the omnipresent and much stronger magnetic fields in everyday life are concerned, we have taken three measures. Firstly, we rely on a diamond-based, spatially resolved sensor that precisely registers the direction from which the signals are coming. Secondly, we use a noise canceling solution, as is also used for headphones. It filters out all signals that do not come from the direction of the muscle or the other signal source in question. Thirdly, we have integrated amplifiers—so-called flow concentrators—which, like a funnel, bundle and amplify all signals from the right direction. With these three measures, we enable the highly sensitive quantum magnetic field sensors to be used under everyday conditions.

Q.ANT recently presented a photonic chip based on thin lithium niobate layers. Where do you see the key advantages of integrated photonics?

First and foremost, I can talk about what we have achieved with the chip, for which we are currently receiving huge recognition—also internationally. We have built a logic processor that can be used as an accelerator and efficiency driver for many computing processes in the field of artificial intelligence (AI). By working optically instead of using electricity, it is much more power-efficient than conventional chips. It requires up to thirty times less power than the latter. At the same time, it is an analog computing instrument that enables highly complex mathematical functions to be calculated in the smallest of spaces. Moore’s law is approaching physical limits: Thanks to EUV exposure, chip structuring has already reached the single-digit nanometer range. It won’t go much further. In addition, the first operators of supercomputer centers (hyperscalers) are building their own power plants to cover their immense energy requirements. Power consumption is so high that it is jeopardizing business models and also putting undue strain on our planet’s resources. If I calculate using light instead of electricity, I need up to 30 times less electricity, so I save on operating costs in a big way. According to market analysts, we are the first company in the world to have succeeded in creating such an analog optical calculator.

The breakthrough of artificial intelligence is rapidly pushing up the electricity requirements of data centers. That can limit integrated photonics. Do you see other applications for your chip technology?

I see this above all in the AI inference, in other words, the chain that is triggered today by a request to ChatGPT. A single request to ChatGPT 4 costs 30 cents. That’s off the charts. More efficient approaches are needed. The thirtyfold reduction in power consumption of our chip is one of them. We have, however, also been able to show that our analog technology leads to huge efficiency gains when training neural networks. For complex problems, users need up to 50 times fewer training parameters to teach the AI solutions. When we recently presented our solution in Atlanta, we had ad-hoc invitations to the hyperscalers on the West Coast. They immediately understood the potential of our solution. Many pieces of the puzzle currently fit together: The mathematical algorithms in the AI sector are predestined for analog computing, but require a huge amount of electricity. And now we come along with the first full-function analog computer that is very energy efficient.

Are photonic computing and purely optical data transmission an option for replacing conventional copper cables and data buses in vehicles and aircraft?

Without being able to look into the crystal ball, I will attempt to give an answer. I think that photonic data processing will move in waves from the world of data centers into the everyday world. The first customers for this are clearly hyperscalers with their high-performance data centers. In a second wave, companies will adopt the technology in order to overcome the existing bottleneck in data transmission. If they have computing power available on site and are networked, transmission-related runtime problems will decrease, and they will benefit from the efficiency of optical data processing. As the technology matures, applications with harsher environmental conditions will then also become conceivable. However, the demands on the chips and optical data transmission in vehicles and aircraft are so high that it takes a lot of engineering to meet them. Due to the efficiency gains, it is nevertheless definitely conceivable that manufacturers will go down this route. However, this requires not only new network architectures with optical cables, but also a complete ecosystem with new suppliers and appropriately qualified specialists. But since the advantages are so great, the industry will use them sooner or later.

To what extent are there cross-connections and synergies in your company between your product lines in photonic computing and in quantum sensing?

There are fantastic synergies, and each of the two use cases is very strong in its own right. I would like to consciously experience how we marry the two technologies together. In conjunction with our highly sensitive human-machine interfaces, high-performance computing can enable direct, non-invasive post-processing of our thoughts directly on the human body. Even if that makes many people flinch at first, it opens up whole new possibilities for us. I can switch the light on and off or dim it with my thoughts, or write texts on a train journey without typing. I can create and use a protected connection between my thoughts and the digital world—if I want to, and if it makes my everyday life easier.

Quantum technologies address future markets. Why is it important to showcase and network in the here and now at trade fairs like World of Quantum?

Quantum technologies are not an economically strong market in their own right today, nor will they be in the future. Nobody buys a product just because it contains quantum technology. It’s about the added value that it can achieve. Messe München has understood that and, with World of Quantum, is very consciously taking the path of embedding quantum technologies. Suppliers and enablers of quantum technologies meet at Laser World of Photonics, which takes place at the same time. Many user industries are represented next door in the automatica halls. Quantum only works in the context of the applications and supplier industry. World of Quantum is the platform that provides this context, where players along the process chain can exchange views and develop joint ideas. We use it to show what is possible with quantum technologies, in order to bring the application ideas in the future into line with the specific requirements of the markets and convince suppliers from the photonics industry that it is worth believing in quantum technologies. As enablers, they often have to invest up front. They will only do that if they recognize the market potential. That is the core around which an exciting ecosystem is currently forming with exhibitors and visitors from all three trade fairs.