Semiconductor Industry Forecast Seminar 2013London, 22 January

Citation

(2013), "Semiconductor Industry Forecast Seminar 2013London, 22 January", Microelectronics International, Vol. 30 No. 2. https://doi.org/10.1108/mi.2013.21830baa.015

Semiconductor Industry Forecast Seminar 2013London, 22 January

Article Type: Conferences and exhibitions From: Microelectronics International, Volume 30, Issue 2

To those of the industry who arrived in London on board snow-splattered planes and trains, a warm welcome came from Malcolm Penn, CEO of Future Horizons, who is now celebrating his half-century in the industry. The need for clarity of vision in this industry can never be underestimated, the lead-times are too long for that, so Future Horizons have been publishing up-dated six-monthly reports which are a reliable, and highly accurate, source of such foresight.

Malcolm began by looking at the global economy, the fundamental driver for everything. There was 3.8 per cent GDP growth in 2011 in spite of all sorts of disasters, and even 2012, with its myriad crises, showed 3.4 per cent GDP growth. The Euro crisis continues, and shows no signs of abating, and this has made matters worse. Pessimism pretty much rules OK, with world leaders fudging the major issues facing their countries, many of whom are mired in colossal debt. The countries with the major debts have what is known as advanced economies, but it is those countries with the emerging economies that show the best growth trends – 5.8 per cent.

The absence of any confidence in markets is the key factor, but the growth rate trends are more into the 3-4 per cent range than those heady 5 per cent numbers we were achieving. From now on it is going to be flat – interest rates will remain flat until 2016, and the Euro will weaken against the US$, with the £ depreciating against the Euro at the same time. The debt crisis in Europe is not going to go away until something radical happens, but with the USA now looking away to other world regions, the economic situation in the EC will not improve, in fact the only rise will be in the level of taxation. The elected leadership in the USA is not dissimilar to the unelected leadership in Europe – both are dysfunctional.

Companies will be struggling to survive this year, and if there is no revenue, there is little hope. Semiconductor companies are not immune, and some of the better-known names are at risk. With confidence at zero, then nothing much happens and people will sit on their money; at the last count it was estimated that business is sitting on US$ 4 trillion.

Investment will only take place in a bullish market. There are of course opportunities in a recession.

Demand is flat, but it is still there, and we are playing games with inventory, which is easy to reduce but slow to replenish. Fab capacity is fixed by decisions taken in 2012, and will still be fixed by 2014, it is a fact of life that from beginning to end it takes three years to bring a new fab on stream. CapEx trend is also flat, which indicates now growth in real terms. Average selling prices have also remained constant; there have been no major changes for some years. However, whilst the individual facts look gloomy, the industry is not in fact in bad shape.

So, what happened in 2012? Malcolm thought that we would be on 8 per cent growth, but outside influences brought the reality down to −2.4 per cent! Confidence was eroded, one fudge at a time.

In 2013, Malcolm thinks that we will see the market growing by 7.9 per cent, with the growth coming at the end of the year. Who knows, the industry may yet break the $300 billion barrier. If 2013 recovers sufficiently, then the market could be worth $500 billion by 2016, as by 2014-2015 it will be starved of capacity.

Malcolm dwelt on the SMART Project that he had been working on for the EC, not without considerable resistance from the European semiconductor industry body, (notably the EECA saying that manufacturing is not important!) relating to 450 mm processing, which will now happen, in a fully automated facility at IMEC, for which the clean room is under construction. Hitherto no one knows who is actually going to make the wafers, but he is confident that a decision on this will be taken soon. Roll-out will be in 2018, which is pretty much as predicted, once the equipment has been designed, manufactured, and, ideally, paid for. Early adopters of this technology will be Intel, TSMC, Samsung, Globalfoundries, Toshiba, amongst others and once 450 mm matures, 300 mm node development will inevitably stop. 450 mm will affect all existing fabs. As the cost structure will annihilate anyone based on 300 mm. Europe will continue as a key contributor to 450 mm technology, as support for IMEC and ASML has been announced By Belgium and Dutch Parliaments, with ASML as the key litho provider, and IMEC the Key R&D provider for processing, materials and transistor structures. Other European firms such as Aixtron, Oxford Instruments, Parts of Edwards Vacuum, LPE Epitaxial, LAM Germany, MW Group and IDC are world leaders in their fields. It is a shame that, hitherto, ST, NXP and Infineon are so unenthusiastic about 450 mm and there is a major political headache developing within the EC on who does what, where, and who picks up the bill.

Looking at the application markets, it was noted that Europe accounts for 13 per cent of semiconductor sales revenues worldwide. In the PC world, Windows XP is still the best system, it works, which is more than can be said for Windows 8, and there really has been little exciting software development since the excellent Netscape was around. Microsoft makes software boring, and user experience is really much too frustrating. Lenovo are now the world’s biggest computer company, concluded Malcolm.

In the mobile ‘phone market, the once mighty Nokia has been on the ropes, and it remains to be seen if Windows Phone 8 can save it. It makes a Blackberry look bad, and Lenovo are joining the fray with their new Smartphone.

Tablets have created a market which never existed. 125 million units were sold last year, growing at 15 per cent per annum.

The shopping list of what is needed for the future included sensors, sensors, and more sensors; lab-on-a-chip analysis; intelligent power control; optical lasers and UV cameras; 5-10 GHz ultra-fast Intel-compatible core processors, and there are seemingly endless possibilities for MEMS. The work being done in the European Consortia Projects (MEDEA and CATRENE) is worthy of a long look, and a final request was for 3D printing (manufacturing) plus robotics, energy, medical and automotive.

The current business model is broken – there is a silicon tsunami on the way, for some good reasons which include the 450 mm transition; the impact of new transistors on chip design and production; advanced lithography (EUV, quad patterning, e-beam, and others); the impact of sub-20 nm nodes on pure play foundry model;. Transistor variability and other “Physics” issues; IDM Foundry new market entrants such as Samsung, Intel and others; the dominance of TSMC; the return to vertical integration, both hard (Samsung) and virtual (Apple). IC production and chip design in the post-CMOS era; the IPhone five type cutbacks and their impact on the supply chain; and the replacement of the current adversarial inter-company (buyer/supplier) relationships, by the more professional and effective airbus industry model.

Thanking everyone one for attending, Malcolm reminded the delegates that the International Electronics Forum will be held on 2-4 October this year, but the venue has yet to be finally decided upon, but will be announced soon. In effect the venue comes secondary to the main event, but Malcolm seems to have a knack of choosing some most agreeable places to visit.