The four-day Esprit ’90 exhibition opened in Brussels on Monday and, with facilities for the world media provided for the very first time, it is making the admirable attempt to show exactly what has been achieved with all that European public money. The Esprit programme, which co-ordinates hundreds of projects whose overall aim is to make the European high technology industry competitive with those in the US and Far East, is set to cost a further 5,700m ECUs (equivalent to $4,104m) up to 1994. Increasingly, Esprit has come under fire both inside and outside Brussels for failing to use this money effectively, and the critics have even started suggesting that the guiding policies of Esprit have in fact done much to retard Europe’s overall information technology competitiveness. In such a climate, the tacit goal of Esprit ’90 is clearly to vindicate the organisation’s policy-making through the achievements of individual exhibitors.
Esprit funds development of Cathedral silicon compiler, US to get benefits
For the last seven years, the Leuven-based Interuniversitar Micro-Elektronika Centrum has been leading one of the most successful Esprit-funded projects in the field of microelectronics. Indeed, if Esprit is all about catching up with the US and the Far East, this home-grown Belgian project is one where the Europeans have actually stolen a lead on the rest of the world. Partnered by Alcatel, Philips, the Ruhr Universitaet of Bochum, Siemens and EDC in Belgium, Interuniversitar’s aim was to develop a silicon compiler capable of significantly speeding design of application specific integrated circuits and very large scale circuits. The result of this project was the prototype computer-aided design compiler Cathedral, which has since been taken on by Philips and vigorously developed into a robust more market-oriented product, the Pyramid compiler. Talking at the Cathedral stand at Esprit, Interuniversitar engineer Hans de Keulenaer reckons that ASIC design for most applications normally takes several man decades and that the result of the project was to reduce the timescale by a factor of between 10 and 20. Cathedral has already been used to develop the circuit for a prototype all-digital graphic equaliser for a Philips’ home stereo system, and for a compact disk application that merges sound, video, text and graphics on the same digital player. It is also being used in the European effort on high definition television. De Keulenaer reckons the Cathedral project has highlighted the need to recognise the development stage within Esprit-funded efforts: previously, Esprit gave a grant for the research phase, and expected the companies involved to bear the rest of the costs involved in getting the product to market. The product development stage was completely ignored. De Keulenaer states: we realised that there was a Grand Canyon between research and the commercialisation of Cathedral. One of the merits of the European Commission is that at last they have seen that this part is necessary and now are financing it. As for the future of Cathedral, the immediate rewards of the seven-year effort seem likely to go outside of Europe as project member EDC is owned by the US company Mentor Graphics, and Mentor has been quick to put itself forward as the natural distributor of the product, which will become available as part of a computer-aided design software package firstly on Apollo, then on Sun workstations some time in 1991.
