back to article Europe to spend €1bn on supercomputers and big data infrastructure

The European Commission’s decided to throw €486 million at high performance computers. The Commission expects that member states will chip in another €500m-plus for a total spend of over €1 billion by the year 2020. All that moolah will be spent on at least four supercomputers, research efforts to develop their hardware and …

  1. Nick Kew

    Why?

    If you spend serious money on a 'puter, you do so for a reason. The reasons stated here seem a bit nebulous, and better suited to the private sector (and up to a point, public-sector consumers like Big Science institutes) than to governments. It might make some sense if it takes the form of a budget for Big Computing projects from the likes of CERN and ESA, but as described in your piece it smells a bit of MeToo-ism.

    And boffins go offshore because they land an opportunity somewhere else. That's two-way traffic, of course. As for the location of actual computer resources? Dammit, I worked on a distant computer in my first job after graduating. If I could do that back in 1983 (with a teletype terminal with no screen, just a paper feed and printing that was fast but also full of line noise), how much more so in an era when we take the 'net for granted?

    1. Voland's right hand Silver badge

      Re: Why?

      Money is like Manure, you need to spread it to make new things grow.

      There is a sh***load of non-project associated university Supers in the USA and China. Most of them got a large chunk of state funding. Someone finally noticed.

      If the new 4 supers are at existing Unis this is not a bad idea. They will recoup themselves once you take 2nd and 3rd order effects into account (educational use, research projects that would otherwise not get compute time, etc).

    2. ToddRundgrensUtopia

      Re: Why?

      The large scale machines whether in the UK, HecTor, EU, (Paderborn), or USA research labs are doing fundamental research, in chemistry, material science, CFD, structural analysis and bioscience, which from time to time delivers something useful. They are also used by a few clever academics who have written code good enough to be able to make use of them. In my opinion well worth the investment

    3. Anonymous Coward
      Anonymous Coward

      Re: Why?

      @Nick Kew - I agree. The stated reasons are nebulous and there's no indication for whom they are being bought - are they're being bought for an identified need, for use by an identified organisation, or are they buying it on the basis of 'if we buy it, they will come?

      It doesn't seem at all clear to me.

    4. anothercynic Silver badge

      Re: Why?

      The EC press release might be a bit woolly, but the ETP4HPC group is anything but. There are several large computing centres (like CSC in Finland, Jülich and LRZ in Germany, the Hartree Centre/STFC, EPCC in Edinburgh and Queens in Belfast, GRnet, SURFsara, TU Delft, ...) involved, along with the usual HPC vendors. This sounds like an alternative to the EOSC project, although EOSC is specifically science-focused with better utilisation of the existing science infrastructures like EUDAT, EGI, PRACE et al, whereas this seems to plan the next hardware generation of said infrastructures. I shall prod some of my colleagues in the HPC game to check if they're aware of this grouping and what they can tell me!

  2. Ben1892

    It's probably a good thing to be late to the party, what with Intel design flaws and everything ( or do supercomputers not worry about things like meltdown? - genuine question!)

    1. ToddRundgrensUtopia

      God I hope they do.

      This european angle though suggests ARM cores instead of x86......maybe.

  3. LaFin

    42

    Will the "10*17 calculations per second" machines be designing the inter-workings of the "10*18 calculations per second?" machines by any chance?

  4. RobertLongshaft

    I for one, cannot believe El Reg didn't get in a sly dig about Brexit in this article. Did you not see the opportunity? Britain to miss out on European super computer?

    Of course this is typical of the EU. Get funding from the taxpayer to build something with little to no value for the government and then lease it our to their corporate friends for a fractions of the market cost.

    It's win win, unless of course you are an EU taxpayer footing the bill for the cultural marxist agenda of the European Union.

    1. Anonymous Coward
      Anonymous Coward

      EU funding an EU project. Why do you still care?

      1. Daniel von Asmuth
        FAIL

        IT made in EU?

        "EU funding an EU project. Why do you still care?"

        Because my country is not exiting the ever more expensive EU for years to come. Never mind the member states having their own HPC projects.

    2. Korev Silver badge
      Boffin

      CSCS (Swiss Supercomputing Centre) is a major PRACE supercomputing centre. Hopefully the UK will continue to participate in European science like the Swiss do.

    3. stephanh

      The Brexit angle would be the question if ARM is still politically possible as CPU architecture.

      From a technical point I would say an ARM-based supercomputer would be interesting.

      1. Korev Silver badge
        Boffin

        There are a few on the way eg Isambard in Bristol, Mont Blanc 3 in Barcelona etc.

  5. Doug Reed

    EU Computer

    They most likely spent less on it than JC Juncker spends on booze and they maybe need it for calculating financial losses or bribes.

  6. Anonymous Coward
    Anonymous Coward

    So many whinging and moaning brexiters in here: Why do they need a supercomputer with muh NHS money.

    Grow up please.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like