So what's the way out of the semiconductor industry's problems? Finding renewed growth and better profitability was very much on the agenda at the opening day of the Future Horizons' International Electronic Forum 2009 in
The chip market is still the same size it was in 2000, employment in the industry has declined 36% between 2001 and 2009, and profitability is on the wane.
TSMC's European President Maria Marced pointed out that: "The weighted average of the profit margins of the major semiconductor OEMs (excluding memory, fabless and foundry) declined from 21% in 2005, to 17% in 2006, to 16% in 200 and to 15% in 2008.
"It has to be made more profitable", said Marced, "and it can only be done by collaboration."
Marced argued that collaboration reduces waste and shares investment, while individual efforts lead to redundant initiatives and heavier investment.
Agreeing with Marced that collaboration is the key was Jean-Marc Chery, executive vice president and chief technology officer of STMicroelectronics.
"We leverage our internal capabilities and multiple external sources to deliver to our business units the best competitive supply price", said Chery.
Chery instanced external sources like: European collaborative programmes like Nanotech 2012, ST's participation in the International Semiconductor Development Alliance (ISDA), CEA/LET, universities in France,
ISDA allows design compatibility between IBM's technology development fab at
Ronald Collett, CEO of Numetrics Management Systems agreed: "Deep partnerships distinguish the R&D leaders", he said, "R&D must deliver the ROI".
Alain Dutheil, CEO of ST-Ericsson, pointed out that, of 8,000 employees in his company, 85% were engaged in R&D.
He suggested another approach to solving the industry's problems: increased broadband penetration is the stimulant the world economy needs, stated Dutheil, quoting a World Bank finding that every ten per cent increase in broadband penetration delivers a 1.2% increase of a country's GDP.
In addition to economic benefits broadband delivers, said Dutheil: "More responsive governments which we need, especially in
Average global broadband penetration is 24.7%.
But the most traditional semiconductor answer to the semiconductor industry's problems came from Young Sohn, CEO of Inphi: "You need to be in the right place going after the right space", said Young, "you have to pick the right space to grow and to bet your company around."
Inphi is growing at 45% a year by making said Young, very high-speed analogue components aimed at Cloud computing i.e. server manufacturers and server farms where customers demand the fastest possible delivery of their information.
Young gave commercial examples of the need for speed.
. Amazon says every 100ms of latency costs them 1% in sales;
' Google says an extra 500ms in search page generation time dropped traffic by 20%;
. Tabb Group says 'a broker could lose $4m per millisecond if their electronic trading platform is 5ms behind the competition'.
"Chery instanced external sources like: European collaborative programmes like Nanotech 2012, ST's participation in the International Semiconductor Development Alliance (ISDA), CEA/LET, universities in France, Italy and the USA and research institutes in Singapore."
- we (the UK) do seem to lag behind our EU peers in this respect, when it comes to collaboration or R&D funding the Germans & French in particular beat us hands down. We are too fragmented, lack focus and direction. Government bodies who should promote collaboration and joint funding seem too busy keeping themselves in a job! There are successful 'cliques' however the barriers to 'getting in' are substantial and there's little help out there. We diminish our competitiveness, limit our opportunities, hurt ourselves. Unfortunately the government does not seem to want to listen. (Sorry for the seemingly political rant, wasn't meant to be when I started typing!)
Is this simply supply-and-demand? With what used to be high profit margins, everybody wants a piece of the pie. The sector is saturating and over-supplied, it is only natural that profit margins waned. With increased competition, surely, the consumers benefit? I just hope this "collaboration" do not result in chips companies forming a Cartel.
Gary you are so right. It's been a bone of contention for as long as I can remember that the UK government has put barriers in the way of UK companies joining in European R&D programmes. I've talked to UK government people about this, and they always say they're trying to get it changed, but nothing happens. The Europeans are pretty fed up and bemused by the UK government's attitude, and can't get it to change. The Europeans would like us in there. Mind you, Andrew Rickman of Bookham, got £160K out of Brussels for some early R&D work. It can be done - but don't expect any help from our government.
Outsider, I don't think it's got anything to do with forming a cartel, more to do with the fact that the market has not grown in value for a decade, but in terms of volume it's grown 20X. So it's running like hell to get less and less return. That can't go on for ever.
Everyone should stop talking about speed and start taking about value to end users. The only new game in town is energy reduction. More functionality usually means trading power consumption. In the end equipment vendors can only deliver dramatic energy savings at the chip level. Lets stop taling about speed and start seriously talking about power reduction at the silicon level.
I am a telecoms professional in R&D and thats where we are focusing our innovation effort!
The only driver for an American family to change out their old fridge is if they can save money on energy. The appliance vendors are scenting an opportunity, what about other industries?
Guys, I hate to point this out but there are still successful emerging semiconductor companies. Mediatek is the obvious example of insane growth with high profit margins.
MTK have successfully attacked the GSM chipset market and grown from 2% to over 20% TAM in just two years. During the same time period TI, Moto, NXP.... achieved the opposite result. I think it goes without saying that TI's R&D team outnumbered MTK's by a ratio of about 10:1
IMHO this is just the beginning, But collaboration is not the answer because it just creates bloated engineering teams and trades product focus for politics.