Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Data storage tapes are stored at the National Scientific Computing Center (NersC) at the Lawrence Berkeley National Laboratory, which will house a US supercomputer, which will work on the future NVIDIA Vera Rubin CHIPS, Berkeley, California, USA, May 29, 2025.
Orbegozo Manuel | Reuters
In Europe, it aims at Gigawatt factories, trying to strengthen its lag in the artificial intelligence industry and solve the problems of the sector rapidly changing.
Noise around the factories concept that have acquired AI production in recent months, especially as Nvidia CEO Jensen Juan emphasized the importance of infrastructure at the June event. Juan said hello on the new “industrial revolution” at the GTC Conference in Paris, France, and – Note Its firm worked to help countries build incomes that receive AI factories partnership In France, Italy and the UK
For its part, the European Union describes factories as a “dynamic ecosystem” that brings together computing power, data and talent to create AI models and applications.
The block has long been backward than the US and China in the race to increase artificial intelligence. With 27 union members, the region is slower when it comes to agreeing on new legislation. Higher energy costs, delay permit and network in an urgent modernization need can also interfere with events.
Heno Virkunn, Executive Vice -President of the European Technology Commission, said CNBC that the target of the block is to combine high quality data sets, computing opportunities and researchers, all in one place.
“We, for example, have 30% more per capita researchers than in the United States, focused on the II. Also, we have about 7,000 startups (which) develops the II, but the main obstacle to them is that they have a very limited computational power. And so we decided that we put in this very serious infrastructure.”
This is a very big investment because they are four times more powerful when it comes to computing than the biggest AI factories.
Henna Virkunun
Executive Vice President of the European Commission on Technology Sovereignty
“We have everything we need to be competitive in this sector, but at the same time we want to create our technological sovereignty and our competitiveness.”
So far, the EU amounted to 10 billion euros ($ 11.8 billion) to create 13 factories of II and 20 billion euros as a starting point for the gigafctor, which means “the largest state investment in the world”. The block has already received 76 expressions that are of interest to a gigane from 16 member states on 60 sites, Virkunun said.
Virkunn noted that the call for interest in gigafactories was “preferred”, going far beyond the block expectations. However, in order for the factories to make a noticeable supplement to Europe’s computing, much more investment will be required for the private sector to finance expensive infrastructure.
“The Revolution of Intelligence”
The EU describes objects as a “single store” for AI firms. They are designed to display the process that is carried out at industrial factories that turn raw materials into goods and services. With the AI factory, the unprocessed data goes into the entrance, and the Advance AI products are the expected result.
According to Andre Kuknin, analytics of UBS stock research, this is essentially a data center with additional infrastructure associated with how technology is accepted.
“The idea is to create a GPU power (graphic processing), so it is basically to create data centers with graphic processors that can train models and run out … and then create an infrastructure that allows you to make it accessible to SMEs and parties that would not be able to just go and build their own,” the kitchen said.
As an object will be used, it is key to its appointment as AI factory, adds Martin Wilki, Citi research analyst.
“You create a platform, having these chips that own a mad level of computing power,” he said. “And if you attach it to the net that is able to get the power to actually use them to full power, the world is in your feet. You have such a huge ability to do something, but what success it will be determined by what you use.”
Telecommunications firm Telenore Already studying possible cases of use of such facilities with the launch of the AI factory in Norway last November. Currently, the company has a small cluster of graphic processors up and running, because it is supposed to check the market before scaling.
Chief Innovation Director Telenor and CEO of AI Kaaren Hilsen Factory and EVP Infrastructure Jannicke Hillland before Nvidia’s counter AI firm
Telenore
“The journey began with Faith – Nvidia had the belief that each country needs to produce its own exploration,” said Telenor Chief Innovation Director and AI Kaaren Hilsen Factory.
Hillsen stressed that data sovereignty is key. “If you want to use AI for innovation and make a business more effective, then you are potentially putting into these AI -important and sensitive business information,” she said.
The company works with Babelspeak, which Hilsen called the Norwegian version of Chatgpt. Technology translates sensitive dialogues, such as his pilot with a border police that cannot use public translation services.
We feel the “intelligence revolution” that resulted in “AI sovereign factories can help promote society,” Hillsen said.
Investing per billion euro
Virkunun said that the first factory of the region would work in the coming weeks, and one of the biggest projects that started in Munich, Germany in the early days of September. This is a completely different story for gigophaktari.
“This is a very big investment because they are four times more powerful when it comes to computing than the biggest AI factories, and it means billions of investments. Each requires three to five billion (euro) in investments,” the commissioner said, adding that BLOC will look to create a consortium.
Bertin Martens, a senior research associate of Bruegel, questioned why such investments needed to subsidize public funds.
“We do not yet know how much private investments have been offered as a supplement to the subsidy of taxpayers, and what potential and how big these factories are. It is still very unclear at this stage, so it is very difficult to say how much it will add in terms of computing,” he said.
Electricity consumption is also a key issue. Martens noted that the construction of AI Gigafactory can take one -two years – but the creation of electricity production of this size requires much longer.
“If you want to create a modern gigan factory with hundreds of thousands of Nvidia chips, you need to count on electricity consumption at least one gigawat for one of these factories. Does Europe have power grids in all these countries to create these factories that remain notice … It will require great investment in regeneration.”
UBS predicts that The current Global Data Processing Center 85 GW will double with -in -the -air demand. Based on the EU EU-EU and EU EUR investment for each factory to launch 100,000 advanced processors, UBS estimates that each factory can be about 100-150 MW with a total capacity for all facilities about 1.5-2 GW.
This can add about 15% to the total power of Europe – significant incentives, even compared to the US, which currently owns about a third of global power, according to data.
Following the announcement of the EU-SSH trading base, EU Head Ursula von der Leyen – Note On Sunday, we, ah -chips, will help eating Gigafactory AI BLOC, trying to help states “maintain their technological preference”.
“It can be argued that this is relatively easy, provided you have money. Buy chips from Nvidia relatively easy and create these hardware factories, but make it running and make it economically viable – a completely different issue,” Martens CNBC said.
He said the EU probably should start from a smaller scale, as the region cannot immediately build its border models in the II costs.
“I think that Europe can gradually enhance its infrastructure and business -out around AI to reach this stage, but it will not happen right away,” Martens said.
