The European Commission has issued the first tentative proposals intended to create a European high performance computer industry capable of competing with established US and Japanese leaders in the field such as Cray Research Inc and NEC Corp. The newly-formed EC working group on high performance computing headed by Director of the Centre for European Research into Nucleonics, Professor Carlo Rubbia, has woken up to the fact that while Japan and the US are currently involved in vast investment programmes in the field of supercomputing, Europe is almost totally absent as a supplier of its own market for high performance computing – estimated to be around 30% of the world total – and suffers from a critical shortage of skilled engineers and scientists in this field. In the working group’s first report, Professor Rubbia details five crucial areas of development: most urgently needed is software capable of exploiting the advantages of high performance computers and their interactive real-time use; simultaneously, Europe should make itself competitive in the design and production of advanced machines and all the equipment required for a leading edge computing environment; these developments would have to be accompanied, the report continues, by the creation of a pan-European high-speed communications network to stimulate the most creative use of the most powerful machines available. Finally, Professor Rubbia called for the mobilisation of industry, universities and large research laboratories together with a concertede training effort to alleviate the chronic skills deficit. Further details as to how these targets could be achieved were sketchy: as a rough figure, it was suggested that, by 1995, an annual investment of $1,410m, from both public and private sources, would be required to fund such an effort; the working group could not, however, put a figure on how much the Commission itself was prepared to throw into the hat. The report itself does not overtly recommend which technology should be adopted as part of the project; Professor Rubbia indicated, however, that massively powerful computing should play a central part, and seemed to reject other technologies such as neural network computers; existing Esprit efforts, such as the Supernode supermini project, were expected to be of value in the early phases of the initiative.
