•  I
• a · A  I
• Accesibility  I
• Map  I
• Search  I
• Staff directory

• ::  Iniciar sessió :: 

The researcher of the Universitat Politècnica de València (UPV), belonging to the Photonics Research Labs-iTEAM, José Capmany Francoy, has obtained an ERC Advanced Grant - the most prestigious grant of the European Research Council - endowed with 2.5 million euros for the development of the ANBIT project.

The project revolves around this idea: to create a new computing model matched to the technological characteristics of integrated photonics, providing an alternative and complementary solution to current digital electronic computing.

"If we are successful, we will be able to develop photonic chips that complement current electronic chips, solving the limitations emerged from the death of Moore's law and generating a paradigm shift in our information society," says José Capmany.

The UPV researcher explains that the cornerstone of this new computation theory, termed analog photonic computing, is a new unit of information: the analog bit (or anbit for short). The anbit is a two-dimensional analog signal that exploits the physical principle of optical interference to perform computational operations.

"This basic idea leads us to enunciate the slogan "Analog is the new digital", which we want to contribute to become the paradigm of the new information society of the 21st century", adds the researcher of the PRL-iTEAM of the UPV.

Second Advanced Grant in six years

The European Research Council has awarded 218 Advanced Grants to prominent European research leaders, 16 of them in Spain. In Capmany's case, it should also be noted that this is the second time he has obtained this prestigious recognition, having also received it in 2017 for developing the UMWP-Chip project.

"For me, receiving this new ERC is an enormous privilege and a source of pride that recognises the excellent work that our group has been doing in recent years, which has materialised not only in publications of the highest level but also in the creation of several spin-offs that are bringing to market concepts and products based on the knowledge that we have generated together here at the UPV," Capmany emphasises.

The team coordinated by the UPV researcher will work on this project for the next five years.

Digital electronic computing: the basis of our lives... but with limitations

Nowadays, computing is indispensable in our lives. Mobile phones, tablets, cars, trains, planes, smartwatches and household appliances make continuous use of it. "For example, something as everyday as a washing machine needs to use computing to know how to interpret and execute the washing program set by the user. It is therefore clear that computing is one of the cornerstones on which today's society is built. However, the current computing landscape, based on digital electronics, presents some limitations, which are the ones we want to overcome with our proposal, taking a step beyond the solutions proposed with quantum and neuromorphic computing,' explains José Capmany.

Among the limitations of digital computing, the PRL-iTEAM explains that it takes a long time to solve problems with multiple data that must be managed simultaneously. This type of mathematical problem is found in applications such as artificial intelligence, autonomous cars, robotics, medical imaging, drug development and quantum applications.

On the other hand, electronic technology is trying to overcome the mathematical limitations of digital computing by integrating more transistors into chips (the basic device of electronics) to give them more computational speed. "However, the ability to double the number of transistors on a chip every two years, known as Moore's law, has come to an end. As a result, the electronics industry will be less and less able to increase the computing speeds required by different applications on chips," says Capmany.

In the face of these limitations, alternative mathematical models that have been proposed are quantum computing - based on the laws of quantum mechanics - and neuromorphic computing – a mathematical model similar to the nervous system and the human brain. "However, the experimental development of both computing theories has been evolving very slowly for decades. A perfect mathematical-technological marriage has not yet been found that coexists with and complements digital electronic computing to solve tasks where the latter is inefficient. And this is what we will try to solve with the ANBIT project,' the researcher stresses.

Analog photonic computing, the solution

According to the PRL-iTEAM team at the UPV, the problem lies in the fact that, historically, mathematical models have been developed first and then the technology has been developed to be matched to mathematics.

"What if instead of developing the technology to be matched to mathematics, we do it the other way around? This is our proposal with ANBIT, to create from the beginning a mathematical model of computation matched to a commercial technology that complements and coexists with electronics. The candidate technology is photonics, for which there is no specific computational model developed. This is the basis of our work for the next five years, a ground-breaking proposal with which, if we are successful, we will contribute to a new revolution in information technologies and society," concludes Capmany.

News UPV TV
Download File