Rail


Is there a need to re-invent integrated circuits?


By Gillian Ewers, PragmatIC


have seen significant consolidation in the semiconductor market, which shows no sign of abating . For smaller companies there is the possibility of using a multi-project wafer service to share the cost, but then the designers must hit specific release dates that only happen infrequently (e.g. SMIC advertises every 2 or 3 months ); get the design wrong and time to market (TTM) can be brutally hit.


The integration path which brings these incredible processors and amazingly compact memory chips will continue to evolve and deliver more appealing, attractive and highly intelligent smart devices. But it does leave an opportunity in its wake for radically


I


still feel young, I am always surprised when I look in the mirror and I see a few grey hairs, well perhaps more than a few if I am honest! But when I look at how things have changed since I took my electronics A level in the early 1980s, it seems like that was light years ago. I can determine who were my degree contemporaries when I ask if they remember transistor man , immortalised in “The Art of Electronics” by Paul Horowitz and Winfield Hill, first published in 1980 (it did actually cover microprocessors and digital buses as well): The first single chip microprocessor was actually the Intel 4004, conceived by Ted Hoff and Stanley Mazor, its 12mm2 was squeezed into a 16-pin package back in 1971 using 10 micron MOS technology by Federico Faggin. Originally a custom design for Busicom’s 141-PF printing calculator and capable of 92 k instructions per second it was hailed as a new era in integrated electronics and was launched as a general-purpose microprocessor towards the end of that year.


Since then we have seen phenomenal


improvements in silicon chips, in terms of complexity and frequency of operation. The acceleration is difficult to comprehend, but there are some


20 June 2019


amazing comparisons to help us: firstly two Nintendo Entertainment Systems (NES – 1985) had the same computing power as the computer that took Apollo to the moon; stepping forward: the iPhone 5 had 2.7x the processing power of a 1985 Cray 2 supercomputer . Then bringing it right up to date – the iPhone X main processor is more than one hundred times more powerful than the iPhone 5, that improvement was over a period of only 5 years! So surely this means we have reached nirvana for all geeks and the general public alike – immensely powerful appliances that are sleek and easy to use? Well yes, and no – more computing ability generally needs more power, i.e. watts, even if it is bursty; and although we are getting more FLOPs for each of our WATTs, there are no major break throughs on the battery front on the near horizon that promise to be revolutionary, so battery life and the weight of our gadgets will continue to be an issue . Also, these meaty chips are expensive to design, it is estimated that the total design cost for a 5nm process chip is on average around $550m . This high price tag is only affordable by a small number of companies, which is one of the reasons we


Components in Electronics www.cieonline.co.uk


different solutions for applications that do not require high levels of computing muscle. For example, IoT used to encompass things like smart cities: lighting, traffic lights, CCTV, etc., but it has now expanded to be an Internet of Everything, IoE, which brings in use cases for everyday items, for example anticounterfeiting, supply chain integrity and tracking, for pharma, food and beverages, healthcare and also interactive games, document control systems and many more. These applications have more pressing requirements like very low cost, flexibility and robustness. It is this opportunity that we at PragmatIC are targeting our first unique flexible integrated circuits at, we call them FlexICs. We envisage the technology as having many other applications, some of which haven’t even been thought of yet. The way we manufacture our FlexICs supports much faster design iterations than traditional silicon, in the order of days/weeks. Using our FlexLogIC automated fab-in-a-box, teams can benefit from staggered development cycles, maximising design and test resource usage and obtaining real result data within an incredibly short time from the original conception of the idea. With low up-front engineering costs, we are set to re- democratise chip design and we look forward to a new era of exciting, innovative creations. pragmatic.tech

Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52