only search Cheshire Henbury

Cheshire Henbury's website is structured around several sub-sites to accommodate the large amount of content. Please pick a topic of interest from the above menu and begin to explore and learn. Or use the Google Search box to the left.

Cheshire Henbury Company Information Pages

Main Home > Cheshire Henbury Information Home > European Projects Home > PROMISE > Bulletin01

Esprit Project 26390 - Provision of OMI Information Dissemination Service


The PROMISE project aimed to improve the effectiveness of OMI's information dissemination by providing support to help and encourages users and project participants with exploitation of results. The dissemination action included an OMI promotional newsletter, and attendance at appropriate dissemination events. PROMISE collected information, identified opportunities and instigated promotion activities and provided feedback for strategy and planning. PROMISE issued newsletters and bulletins and organised the OMI conference (EMMSEC) that was be held in Florence in November 1997.




Information dissemination, electronic commerce, multimedia, embedded systems.




Cheshire Henbury (UK)


PROMISE Results - EMMSEC Conference

Link to EMMSEC Home Page (Opens in New Window)

PROMISE Results - Project Bulletins

Hardware and OMI

In the late 1980s, there was a continuing decline in the European share of semiconductors, and microprocessors in particular. This was despite considerable innovation and technical developments. JESSI (Joint European Sub micron Silicon Initiative, now MEDEA- (Micro-Electronic Development for European Applications) was tackling the process and manufacturing technology, with companies such as SGS-Thomson, Philips and Alcatel collaborating on base technology e.g.HCM0S5.Technology is no good without products, and 0MI was to form the bridge.

Although aspirations still existed for the European computer industry, the real opportunity was foreseen in embedded systems. Europe had a strong presence in the emerging RISC (Reduced Instruction Set Computer) microprocessors with the indigenous ARM and transputer, and the licensed SPARC and MIPS architectures.

OMI's Strategy

It was clear that densities i.e. the number of transistors in. a given area, would increase. Moore's Law predicting a doubling every 18 months or so had held for 20 Years, but with hindsight, speeds probably grew faster than many of us realistically expected.

In 1990,with 1.2 micron technology, a 32 bit transputer was a 10 x10 mm-chip, but actual CPU was only a small portion - on chip memory, communications and other functions accounted for over 60% of the chip's 300k transistors. In a few years, with millions of transistors, the overall die size would be unable to be reduced because of the number of bond pads (the parts around the edge of the chip where the wires connecting the outside world are attached) would not be decreasing. The big question by the late 1990's would be, "what to do with the silicon real estate?"

For desktop computers, the answer was obvious-the CPU would be improved, and any extra space left would increasingly be used for cache memory. For embedded systems, the only obvious strategy was to put more of the system onto the chip. So reducing external components and cost and increasing reliability.(Modem chips have a reliability that is roughly proportional to the number of pins they have .Reducing from 10 components with 100 pins each to 1 component with 100 pins can give an order of magnitude improvement in reliability).
This is the system-on-a-chip that is common today, but at the time, this was thought to be somewhat visionary and impractical.

System on a chip

And this is where 0MI came in-if a large pad of the system is going to be on one chip, who will design and make it? Previously, the systems integrator would have designed the system from standard chips, and perhaps a few ASICs (Application Specific Integrated Circuits). We now faced a fundamental business, not technical, problem. Europe's strength lies in embedded systems, with a large number of smaller companies involved. The scenario above leads to only large companies being able to participate and compete!

There is a solution: companies can work in partnership with each contributing their expertise. For this to work, designs could no longer be unique every time. Much as we build houses from standard components e.g. bricks and when we need to do this more quickly, from prefabricated panels. Transistors are the electronic equivalent of bricks; now we needed higher level functions AND a common framework so that the components would together.

Applying divide and conquer to the problem, it breaks down into issues of methodology and standard interfaces.


0MI was launched with a standards project to address these issues, and a series of demonstrator projects in which the techniques would be incrementally applied (businesses could not wait for the perfect solution, or they would miss the market opportunities).

So how did the companies respond to the challenge? Together, in the OMI Standards project, Siemens, Philips, ARM, Temic (Matra) and SGS-Thomson (Inmos) developed a common basis for an on-chip bus, for connecting peripherals. This provided a focus for other companies to develop modules, and encouraged more entrants, such as Nordic VLSI, Hyperstone, Mitel (GPS), SDSA etc.


0MI Bulletins are occasional leaflets published by the OMI PROMISE project.

0MI,The Open Microprocessor systems Initiative, is a programme set up by the European Information Technology industry and the European Commission in the framework of Esprit the European strategic Programme for Research and Development in Information Technology.
Let's look at some of the products that have been developed. Accepting that most of these are not directly recognised, as they are buried within something else i.e. the application.

Working in partnership with car manufacturers, SGS-Thomson took the transputer and turned it into a modular 32 bit core, and for the first instantiation, added the necessary functionality to create a cost effective system for GPS (Global Positioning by Satellite) to create in-car navigation systems. SGS-Thomson is purely a semiconductor company, but with a full range of facilities from design through manufacturing to packaging and test.

ARM developed the ARM family of cores. ARM is a small company, founded at the start of 0MI as a spin-off from Acorn, who had developed a 32 bit RISC processor for their own computers. ARM was thus one of the early fabless design companies. Partnerships with semiconductor foundries, as licensees, were an essential part of the business plan. An early example brought ARM and Mitel (formerly GEC Plessey Semiconductors-GPS) together with one of the resulting chips being used in Apple's Newton. Newton was the first hand held PDA (Personal Digital Assistant) to use handwriting recognition, a feature previously impossible without the computing power at low power consumption introduced by the ARM processor. Another example was the partnership with Hagenuk to produce a cheaper ISDN telephone supporting digital services. Yet another was the use of the ARM microprocessor by OptionExist to create the portable FAX, marketed by Fujitsu.

Philips is a large vertically integrated company, including design and manufacturing not only of semiconductors, but a wide range of electrical and electronic products. They took a licensed processor, MIPS, and modularised it for use with the 0MI peripheral bus, allowing system chips to be very quickly developed through a very high degree of re-use. A side effect of high re-use is a lower level of design problems-most of the design is already well proven. With a standard set of modules, it becomes feasible to make progress. n the related area of hardware/software codesign - techniques to assess which parts of the design should be realised with dedicated hardware, and which are more cost effectively implemented in software. This is a good example of the virtuous circle 0MI set out to create-improved design m-use increases productivity, increases reliability and reduces time to market and costs. Some of the resulting chips can be found in mobile phones and PDAs.

Siemens is also a large vertically integrated company, but the semiconductor division operates autonomously. Work done in 0MI led to the 80C167, where a fuzzy logic co-processor was added to an existing 16bit controller and used in control systems for electric vehicles, where there is not a straightforward relationship between the measurable parameters and the system to be controlled. 0ther work has led to the development of Tri-core- a 32 bit processor with some DSP (Digital Signal Processing) characteristics, targeted initially at the automotive market.

Temic (Matra) took the licensed SPARC architecture and developed a family of communication controllers. This has now been spun out into a separate company, Tsqware. These products are typically used in data concentrators ISPs (Internet Service Providers). This highlights a feature of the emerging IP (Intellectual Property - in this case, semiconductor design) business. There are a limited number of people with a deep understanding of a particular type of system. The world market will buy the IP, rather than develop and maintain staff with that expertise.

Integrating processor cores and developing a range of peripherals is not enough to be successful in the silicon systems on a chip. The business case requires them to be cost effective. Generally, for the lower cost mass market, this requires that the chips are put into cheap packages i.e. plastic, which have limited capacity to dissipate heat. Secondly, many of the embedded market opportunities are for new products, especially portable new products. Silicon technology development will evolve ways of reducing the power, but
OMI has pioneered a radical new approach - application of asynchronous logic to real systems. This computing-without-clocks makes use of the fact that the CMOS technology in common use for the last 20 years only consumes power when switching. Most of the time, clock signals are fed to every circuit, but other than the clock signal itself, there are no changes: many signals have low rates of real change. Thus if signals only change when it is meaningful for them to do so, considerable power can be saved, Within OMI, Philips have developed this technique and applied it into a portable CD player and Manchester University are now on their 3rd generation implementation of an ARM microprocessor with this technique.


Some of these are examples of what ESPRIT has helped to achieve - new ideas and businesses grow up around a capability, as in Silicon Valley, California. In Europe, 0MI has, through partnerships, and successes of the companies involved, helped change the thinking of engineers to a belief that new ideas and innovation can succeed here in Europe. This has come about through the transition from a Euro-centric view of business to a global view of the business. Ironically, a number of companies have not survived in the form that they started in, having either ceased to trade, or been taken over. This morality suggests that the market in Europe is now much more driven by the facts of business life, and that ESPRIT is helping to overcome the lack of critical mass in any one country, by reducing the barriers, financial and otherwise, to creating a distributed critical mass. This is essential for a true single market to develop, as it must encompass design and development, not just sales of products.

In this analysis, we have been looking at a few of the hardware successes- however the silicon hardware is just one part of a three, way partnership:




0MI Bulletins are occasional leaflets published by the OMI PROMISE project. 0MI,The Open Microprocessor systems Initiative, is a programme set up by the European Information Technology industry and the European Commission in the framework of Esprit the European strategic Programme for Research and Development in Information Technology.

Whose software makes the world go around

Tooling up for tomorrows world

How smart can you get?

Keeping the motor of the European economy in good condition




Some of Paul T Kidd's Books

Book Covers

Legal Notice: The information posted on the web site is designed to provide accurate and authoritative information in regard to the subject matter covered. Every effort has been made to ensure the accuracy of the information contained in the web site. The information is believed to be correct at the time of publication. Cheshire Henbury cannot however accept any responsibility for the completeness, accuracy and relevance of the information. Information is published with the understanding that publication does not represent the rendering of advice, consulting or other professional services. Specific application in a particular organisation is the sole responsibility of the representatives of that organisation. If expert advice is needed, the services of a competent person should be sought. Please read our terms and conditions (opens in a new window) for use of this web site.

Cheshire Henbury

Address and Phone Details (opens in new window)

Email: Contact form (opens in a new window)

Web address: