The United States and China have a considerable advantage over Europe and Japan in managing the technological ecosystem, and Europe does not have an industry that allows it to have sovereign components to meet its needs. Giuseppe Gagliano's analysis
The European initiative projects in the field of artificial intelligence are fully in line with the Horizon 2020 program launched in 2013 but also with the Horizon Europe program that will take over in 2021, with the aim of financing Union research and innovation. European. The latter – as is well known – has set itself the goal of creating a European high-performance computing infrastructure of world-class.
The heart of the ongoing technological battle at the international level between the various artificial intelligence players is therefore essentially centered on computing power. High-performance computing is in fact necessary to simulate and model complex situations in the industry and research sectors (Machine Learning and Deep Learning, processing large volumes of data to respond to climatic, seismic and energy problems, etc. ).
Among the vectors of high-performance computing, we distinguish supercomputers, also called HPCs for high-performance computing, and quantum computers (almost infinite computing power in minimal times with a low error rate). The latter, which appeared in the 1990s, are still only at the prototype stage. However, the private entity or country that takes command of the HPCs will have significant technological advancement for quantum research.
The challenges of the European Union's control of computing power are to develop its independence and sovereignty in computing.
However, this desire for European sovereignty and independence could be altered by two issues. The first due to the presence of foreign private actors in European initiative projects and the second due to the difficulty of using sovereign components in supercomputers and quantum computers (microprocessors in particular, come mainly from the United States and China).
The absence of a common policy for the European Union through a European "buy act", similar to the American one known as the "Buy American Act" in force since 1933, is at the origin of the European Union's loss of speed in control of its value chain in the face of the quantity of American, Chinese and Japanese components on the market. This lack of vision is certainly a serious obstacle to innovation on the path of strategic independence.
The EuroHPC project with a budget of around 9 billion euros in the period 2018-2033, aims to develop high-performance computing infrastructures. About 4 billion euros come from private European and non-European actors, the rest is financed by the European Commission and the 12 member states of the project. The project board includes private partners such as ETP4HPC (European Technology Platform for High Performance Computing), BDVA (Big Data Value Association) and Prace (Partnership for Advanced Computing in Europe). Well some of the members of this organization represent foreign subjects.
In May 2017, the Vice President Europe and Director of Research at IBM in Zurich made it clear that IBM was very early involved in European research programs alongside American research programs. As such, IBM is one of the founding members and present on the board of directors of the German think tank ETP4HPC. Similarly, Intel has served on the board of directors since its inception, and Fujitsu joined it in 2018. These companies are therefore able to influence research directions based on their own technological advancements.
If we think of Prace, a pan-European international non-profit organization, it provides infrastructure for scientists and researchers in academia. One of the six Prace reception facilities is located in the laboratories of the Bsc (Barcelona Supercomputing Center). In 2015, the CEA indicated in a statement that together with the BSC, they are working together to create innovations in the field of high-performance computing even though the BSC has partnered with manufacturers such as IBM, Microsoft, Intel and Nvidia. Now let us ask ourselves: how can the European Union improve its sector when such partnerships are in place?
Let's now look at BDVA which is a group of industrialists closely linked to the European Commission, born with the aim of developing an innovation ecosystem that will enable the digital transformation of Europe through data and artificial intelligence. In particular present on the board of directors, the vice president of the European research institute of Huawei Technologies in Düsseldorf, the program manager for the European Union of IBM Israel, as well as the project and data protection manager of NEC Laboratoires Europe (Japanese). Again, aren't the interests of these powerful foreign industries in terms of information acquisition and decision influence at odds with the objectives of the European Commission?
The emblematic players in the field of components and information technology, namely Americans, Chinese and Japanese, have long understood the value of participating in the financing of European projects. This is one of the reasons why they have established themselves in Europe through subsidiaries and partnerships with research centers and industrial groups. This we enter has no other purpose than to capture knowledge, plunder technologies and influence decisions according to their interests. In other words, the stakes revolve around the mastery of the ecosystem.
The United States and China have a considerable advantage over Europe and Japan in ecosystem management to the extent that Europe does not have an industry that allows it to have sovereign components to meet its needs. Japan, for its part, appears to play a secondary role with partial control of its environment.
The ability to be autonomous from design to implementation is a real strategic asset. For example, thanks to the increase in world demand for components and supercomputers, China – renowned for being a copying industry thanks to cyber espionage – is able to carry out a veritable "technological plunder".
Two other important technical problems are added to the previous three and concern the optimization of power requirements and energy cooling, corollaries of high performance computing. In fact, the more powerful supercomputers are, the more energy they consume and the more heat they emit. One of the challenges of improving computing power is developing components that require less energy and also consume less energy in terms of cooling. The optimization of the energy efficiency will allow to realize more and more complex and efficient systems.
Ultimately, doesn't the presence of foreign actors working on the development of European high-performance computing run counter to the European Union's desire to achieve its leadership goal?
In the end, isn't Europe, in spite of itself, an area of expansion for foreign appetites?
The European Commission should refocus on a pan-European governance system that is more protectionist than the funding source and components.
This is true technological sovereignty.
This is a machine translation from Italian language of a post published on Start Magazine at the URL https://www.startmag.it/innovazione/tutte-le-sfide-per-la-sovranita-tecnologica-dellue/ on Sat, 17 Oct 2020 05:59:57 +0000.