The model of 'Keynesian economics' plays a massive role. In fact this form of economic model does not require manufacturing for a successful economy and that's why the UK stopped supporting all manufacturing - not just Fabs (blame Margaret Thatcher and John Major). Compare the banking systems (service industry) and you'll find the UK are well ahead of the USA in banking infrastructure.
It's got nothing to do with differing classes, old blighty, or any of this nonsense and most Fabs were actually located in Scotland. The remaining cost competitive Fabs worth talking about are located in Greenock Scotland (National Semi) and Newport Wales (Int'l Rectifier).
Barring a few exceptions ( e,g. Bombardier ) and transient blips like RIM, Canada has always been primarily a source of raw materials rather than engineered products.
The scaling laws that drive semiconductor devices are far more aggressive than for bulk steel. So in spite of some superficial similarities between the two in their commercial trends, the Physics barrier for semiconductors will be hit real soon and this is expected to mess up the current growth of offshore Foundries and the free ride enjoyed so far by the fabless wonders.
The lag time to introduce new technologies between Fabs owned by the leading IDM ( in US ) and large Foundries based in small islands is already increasing. To convince yourself dig up the dirt on the introduction of high-k metal gates. It will only get worse with FinFETs.
If silicon is a new steel the key question is whether it pays for UK (or any other developed country)to be in silicon manufacturing business? The answer seems to be NO. Canada, the country I live in, has largely got out of silicon manufacturing a while back and there are similar examples around the developed world...Kris
I left South Africa to go to Oz and have a UK passport, so you can calibrate my view from the Penal Colony. The UK had a social bill of almost $100 billion (NHS and bludgers), debt of $9 trillion (equal to that of the PIIGS and
Australia), and so much compliance nonsense (also in Oz) that it is almost impossible to run a business. I have yet to see people take sick leave so easily and get big gun drugs for back pain. You want to stop the boats and ex-colony pouring into any country with benefits--easy; make them work. If governments subsidise workers in China and elsewhere, they have created employment while elsewhere they destroyed it. I earn less than a trades person (have M.Sc and thirty years experience), wife is a doctor with the equivalent of the NHS in Oz and still pays a 40% service fee and then also GST on that and gets taxed heavily. Kids pay $24,000 for university because we were "foreign on temporary visas". She is forever dealing with bludgers on Centrelink with compliance issues. Luckily Oz has rocks to sell and only $1 trillion of debt. South Africa -- crime and under $100billion of debt. I went down the Transputer, DEC Alpha, MIPS and PowerPC route and watched Compaq swallow DEC, HP swallow Compaq and then ditch what they paid for. All the big IT companies that went the x86 route thought they could markup high, but the East is competitive and that is where we all shop, even for underpants. Tax as well; anything above 30% is a problem, but if Cameron thinks electronics companies will flood into the UK with 20% VAT, he must be taking something. I watched a speech of his in amazement a few months ago--not sure if he believes it himself. I'm off to retire, hope Wall Street and the imaginary debt cash-out does not plunder my fund too much. Good night.
More-than-Moore has a role to play although much of the chip industry economics is based on the continuous movement of the downward moving staircase.
When that stops the rules of the game will change significantly.
One could argue that the rise of More-than-Moore is partly a response to the exponentially increasing cost of moving down that staircase.
For sure fewer designs are being done at leading-edge geometries and those that are need higher volumes. But there is plenty of value to be created in design at well-established nodes.
So, yes, expect variety -- fabless chip companies, plastic electronics, MEMS, packaging all to have an increased wealth-creating role to play.
The landscape is more dynamic and complex than ever before.
On Intel in Ireland I do not think there was any significant operational cost difference to Scotland. The difference was that Scotland had already attracted a number of inward investment fabs and had "aid fatigue" while Ireland was prepared to do everything it could in terms of support and subsidy to attract Intel.
Add to that the large numbers of excellently educated young people coming from schools and universities in Ireland and Intel was happy to bite. The downside was Intel had to foster the support infrastructure for its fab. The upside was that nobody was going to lure trained staff to across town to a rival chip maker.
Your points are well made. Some UK-owned and some foreign-owned mature, well-amortized fabs do continue to operate in the UK.
However, if we compare the value of the ICs and discrete semiconductor devices produced each year in the UK compared with, say, Taiwan I think we would find a very big difference.
Can NMI you say what is the annual market value of the chips produced in the UK?
And when you say chip manufacturing companies have invested significantly in the UK in the last few years how many billions of dollars would that be?
It is my contention that if the investment is not measured in billions of dollars, while it may matter to the companies concerned and be valuable for keeping their mature fabs going, it is not really significant on a global scale.
That's not pessimistic or bleak. It is just the way it is.
Peter, interesting article. Seems you ruffled the feathers of the great and the good manufacturers of the UK as this is a very pessimistic and bleak opinion and discussion… suggests (ultimately) nobody in the world should do manufacturing! Working for the trade association who represent UK companies within the Microelectronics sector I can honestly say that I am amazed and proud of how well the UK fares in such a competitive landscape. If I may, here is a more balanced and accurate view. The historical events stated in your article do recognise the decline of indigenous fabs but you fail to appreciate where the UK is today as far as manufacturing is concerned. THE UK IS STILL IN FABS….just foreign owned ! There have been a few high profile fab closures in recent years like Atmel, Motorola and NEC, arguably these decisions were not made on fab performance but due to other political and corporate issues. However despite this damning article, despite outrageous energy costs compared with other regions and minimal funding assistance or tax breaks compared to other countries, the UK still produces leading edge products in a wide array of applications in many fabs and importantly who make money. International Rectifier, Seagate Technology, NXP, National Semiconductor and Diodes Zetex for example all have fabs here, all have invested significantly in the last few years to increase capacity and capability of their sites. Many other smaller fabs exist too, producing specialist niche products successfully. These operations should be congratulated on their continued success in spite of the uneven playing field they find themselves in. Why not come along to the NMI Manufacturing Excellence Conference on the 5th of October and see some great of examples of UK innovation and success in fabs.
David Law, Director of Manufacturing Services, NMI. www.nmi.org.uk
Would Peter care to comment about what happens in the US beyond the new Fabs now under construction in my hometown ( Intel ) or in Malta, NY.
Any thoughts on whether " More than Moore" technologies might skew down the cost escalation curves for new Fabs or their break even volumes so as to enable smaller ( with less than 5 billion $ in the bank ) but innovative players once again ?
Blog Make a Frequency Plan Tom Burke 17 comments When designing a printed circuit board, you should develop a frequency plan, something that can be easily overlooked. A frequency plan should be one of your first steps ...