4 ms·
I don't know how that relates to EU projects. I'm not alleging corruption, just stating facts. The flagship projects were promised 1 billion in funding over 10
by orbifold 7y ago
I don't know how that relates to EU projects. I'm not alleging corruption, just stating facts. The flagship projects were promised 1 billion in funding over 10 years. A lot of that money was spend in Switzerland, which isn't even a EU member state. The last funding period will now actually involve 20% in administrative overhead, spend on administrators mostly located in Switzerland. I've seen some of the planning diagrams for future European cloud initiatives especially in the HPC sector. Most of it is just shuffling around funds from one budget to another, that is they relabel existing HPC infrastructure and university compute resources to be now part of a "European Research Cloud" (https://ec.europa.eu/research/openscience/index.cfm?pg=open-science-cloud https://ec.europa.eu/research/openscience/index.cfm?pg=open-...)
Another aspect is that as soon as a university is involved, you get lots of cheap basically black labor by gullible university students. Take https://qt.eu https://qt.eu for example, they promise 1 billion in funding over 10 years and 5000+ researchers. That works out to 20000€ per researcher per year. A PhD student in Germany makes roughly double that, if he has a fully funded position. So this figure relies on:
- Paying people half the normal salary, but expecting them to work full time, which is common practice in European academia.
- Significant additional funds being provided by member states and universities, all the while the European Union has the leverage to call the shots on research direction and funding.
- ivan_gammel 7y agoMy point is that various overhead and mismanagement does not mean that project goals won’t be achieved. If budget is big enough, it may be more effective to work with already existing administrative structure to get things done rather than trying to improve management and optimizing costs.