Esprit ’91 isn’t as open as it claims: uniformed guards turn out non-badge wearers
Organisers of Esprit ’91, which likes to pride itself on its openness and accessibility, were somewhat embarassed to discover that earlier last week the uniformed guards at the entrance of the Palais des Congres had taken to ejecting members of the public who had gained entry without a badge. What the guards should have been doing, explained a despairing public relations representative, was to tell visitors that they needed a badge, but that badges were freely available at the kiosk round the corner.
New liquid crystal technology may reduce the cost of personal computing
The European Community’s Esprit research programme is working on two liquid crystal technologies, still in their infancy, that could hold the key to bringing down the price of portable computers. If active matrix liquid crystal displays, or LCDs, solved some of the problems associated with the old, electronically immobile LCDs, such as the perpetually disappearing arrow, then the low manufacturing yield rates achieved for active matrix LCDs must now really be testing the resolve of Japanese manufacturers like Sharp Corp. Stephen Hawes, project leader at GEC Plc’s GEC Research Active Matrix Manufacturing facility in Wembley, London, estimates that at best only a small percentage of production volume actually comes through as finished goods. The joint GEC-AEG-IMEC project has discovered that the low yield rates of current active matrix liquid crystal displays results from the amorphous silicon used. This requires large transistors, is photo-sensitive and suffers from parasitic capacitance, in that the charge of one pixel affects that of neighbouring ones. Solutions to all these problems can be found, but they complicate the design process. GEC has compared this to results achieved with a newer, polycrystalline twisted-field transistor design. The glass used, according to Hawes, is slightly expensive, though this factor is offset by other features. The polysilicon liquid crystal display is not photo-sensitive, so there is no current leakage; it works with much smaller transistors and, because it runs at a higher temperature, the problem is actually running it a low enough temperature; there is no parasitic capacitance. Even in the pre-pilot production phase, in which the Wembley plant finds itself, says Hawes, a yield rate in excess of 50% could be achieved, which would result in cheaper, European-built liquid crystal displays, and could cause portable computer prices to tumble. There is just one snag. While GEC’s Lord Weinstock tries to sell the technology to manufacturers such as Philips Electronics NV, Hawes estimates that the costs of bringing the present facility up to full production capacity will be essentially similar to those required for integrated circuit production.
Ferroelectric LCDs may break out soon
Ferroelectric liquid crystal displays were thought up just a few years ago and the technology suffers still from problems of temperature which need to be solved before a niche in the bottom end of the LCD market can be identified. The advantage of ferro-electric liquid crystals is that they are cheap and have a simple structure, making it possible to manufacture large, flat-panel displays at low cost. The 50KHz switching speed and bi-stability – so the display doesn’t need to be continuously refreshed, means they can be used in high-contrast, low power-consumption devices. Applications, which could be out soon say the developers, may include cheap colour video or high density, fast shutter arrays.
ASM Lithography wafer stepper can design tiny 16M-bit and 64M-bit DRAMs
Content from our partners
The three-year Deep Ultra-Violet Lithography project, to develop equipment capable of making integrated circuits with features measuring less than half a micron, is being paraded as one of the big successes of Esprit. Due to be completed this month, it has already helped one of the project members, ASM Lithography, of Veldhoven in the Netherlands, to design a wafer stepper that can desig
n 16M-bit and 64M-bit dynamics to these specifications. ASM Lithography, according to one Esprit programme director, is one of a number of companies that have effectively used the Esprit grant as seed money. Now clients for its PAS family of wafer steppers include IBM Corp and Advanced Micro Devices Inc in the US, SGS-Thomson Microelectronics BV and Philips Electronics NV in Europe, and ASM feels justified in describing itself as the market leader in Taiwan. An ASM spokesman hinted that a large deal with a Korean wafer manufacturer was imminent, and claimed that it was actively involved in talks with several Japanese manufacturers. Japan is now the tough but crucial market for ASM to crack, representing 50% of the world’s integrated circuit production, according to estimates, and currently dominated by indigenous suppliers like Canon Inc. But, despite partnerships with Europeans like Philips and SGS-Thomson, ASM has still to convince others of the project’s worth. Managing director Willem Maris has issued a statement complaining that some European semiconductor manufacturers, Siemens AG for example (which, according to ASM, has bought Japanese), were still favouring foreign, but inferior, wafer stepper equipment.
Consortium releases Mermaid Software Development Management Software
The National Computing Centre, the City University of London and Utrecht’s Volmac Software Groep NV have come up with a piece of software which, in an ideal world, could make vapourware a thing of the past. The Mermaid software is designed to give software project leaders an accurate estimate of the time the team is likely to take on a software project. Project members claim its uniqueness lies in the specific criteria it uses to make the calculations. Among other things, it takes into account what the finished software is designed to do, the intelligence the programmer is likely to require, the languages and methodologies being used, and the depth of experience of the design team in this area. The Windows-based program displays project data and its conclusions in different ways, so that a project leader can pinpoint what elements in the project are taking the most time. It could tell you, for example, if a switch to a fourth-generation language from Cobol mid-project would result in any time savings. Volmac’s Jan Groot says Mermaid is currently being beta-tested in a number of sites, typically financial institutions interested in developing complex applications. Volmac reckons that the software will be out in May for UKP14,000.
