Gönner said investing in new technologies, such as artificial intelligence, was vital.
Huge amount of money and resources is already (mis)allocated to LLM-style AI. With very dubious results on productivity, jobs, and clearly bad results for the environment.
It’s surprising to see people wanting to pour even more money into this pit.
It’s important to understand that the meritocracy is a lie and all these goons just vibe their way through repeating what buzzwords seem appropriate.
Is Germany allocating a lot of money to LLM stuff? I see positively the increase in computing power available in the EU: consumption is there, just outsourced from other countries.
I do see positively European projects such as AI factories. Eventually they may not even be used for AI development, but a strong computing infrastructure accessible to researchers and small companies will definitely be useful for the development of European technology.
LLMs require specialized processors, ie TPUs and ASICs. Using GPUs or CPUs would be even less efficients. AI datacenters layout and cooling are specifically designed to accomodate those specialized processors.
My understanding is that repurposing such datacenter for other kind of computing would be expensive. Doing this shortly after building them would mean loosing a signifiant portion of the initial investment. ie would require replacing processors, cooming, and redisigning the overall layout to optimize for another type of processing.
AI factories have CPUs and GPUs. Either way, TPUs can be used for other kinds of calculations as well. And definitely, given enough availability of TPUs, specialized software will be optimized to run on them.
But either way, no this is not the case. I know people who work with LLMs and they mostly use GPUs, even when they work on US computing clusters.
Fair point, GPUs are often used too, and have other uses for HPC workloads.
There is still a major misallocations of resources for AI. This infrastructure is badly needed by big tech for their AI growth plans. No one else need this much.
There’s no good reasons to build at that scale and so fast, for any reason.
There’s a need to lower overall energy and resources usage, and improve efficiency. Building many large datacenters for LLMs go in the wrong direction.
I work in research and I have dire need of GPUs. I do not work on LLMs.
Recently we got granted access to 500 GPUs for 2 months; all of them are spinning - and this is the work of 5 researchers. A normal research HPC institute normally has some 100-200 GPUs. In this sector there are startups which need to execute simulator calculations.
Yes, GPU infrastructure is very much needed and lacking in Europe.
I’m all for GPUs and memory being available to research that benefit public good, within limit. And less of it being available to LLMs.
AI companies are swallowing a large part of memory and some chips production, meaning less availability and higher prices for everyone. This is part of the resources misallocation. https://www.cnbc.com/2026/01/10/micron-ai-memory-shortage-hbm-nvidia-samsung.html
This project is making computing power available to research and small companies.
https://digital-strategy.ec.europa.eu/en/policies/ai-factories
GPUs are now more widely available than ever before.