2013: Trends, Issues, Hopes and Fears

The problem for the semiconductor industry is lack of growth. The semiconductor industry has always relied on growth to fund its next generation of products.

The problem for the semiconductor industry is lack of growth. The semiconductor industry has always relied on growth to fund its next generation of products.

But, as silicon technology approaches its limits, there is less and less reason to buy the new products.

The good old days, when every new node produced higher performance, lower power and greater density, are gone.

Nowadays the industry has to resort to increasingly expensive manufacturing tricks to get to a new node and then the new node doesn’t deliver the significant product improvements which used to be relied on to persuade customers to invest in a new generation of chips.

Which is why the FD-SOI vs Finfet issue is important.

If chips made on planar FD-SOI turn out to be lower power and cheaper to make than chips made on complex 3D finfet processes, then it’s another significant blow to Intel’s faltering aspirations in mobile.

Globalfoundries is installing the FD-SOI process transferred over from STMicroelectronics. If big players like Qualcomm and Broadcom decide to use FD-SOI then Intel’s bet on finfets could look misguided and  costly. While FD-SOI could be the catalyst ST needs for its digital business.

Intel’s leadership issue is mega. No one knows why Otellini left, so no one knows if Intel is looking for a radical shake up.

A 25% drop in share price is probably enough to trigger a change in CEO at a US company, but there are other possible reasons: fallings out with Warren Buffett, Microsoft and Apple, massive over-building of capacity, the running up of big debts, the persistence of rumours of unsavoury marketing practices.

Also, it looks as if Intel has run out of competitiveness. That could be result of the innovation-sapping effects of having a monopoly. The old fizzy days of the Noyce/Moore/Grove troika when Intel was super-confident, bold and persistently innovative are long gone.

Could, one wonders, Intel find an inspirational new leader who can re-establish the Noyce/Moore/Grove culture? IBM, in similar circumstances, did exactly that when it found Lou Gerstner. We should know in May, when the new Intel CEO takes over.

Japan is addressing its problems in a very Japanese-style way. Renesas appears to have been nationalised. Sharp has sold 2.5% of itself to Qualcomm for what seems like a paltry $60m and may sell another 2.5% for another $60m. If Sharp survives, Qualcomm’s got an outstanding bargain. But $120m is not going make much of a dent in Sharp’s $12bn debt mountain and Sharp’s own management has said it may not survive.

Qualcomm has been on a stonking run lately ending 2012 as the third largest chip company in the world. But a shadow hangs over it. As the two largest phone and tablet manufacturers – Samsung and Apple – move to making their own chips, is there enough business out there to sustain Qualcomm’s progress?

And can Nokia, RIM and HTC survive in the face of the Samsung-Apple duopoly?

Then, of course, there’s Apple. Apple shares are down 27% from their peak – about the same level of fall as when Intel announced the departure of CEO Paul Otellini.

$175bn has been wiped off Apple’s market cap as the share price falls from $700 to $500. Sales of iPhone 5 said to be poorer than expected.

Apple needs a big trick – something revolutionary rather than incremental improvements to existing products. The proposed TV project could have been such a thing but there are rumours that the TV project has been put back a year.

Rumours about Apple producing a watch hardly seem like something which can resurrect the company. It beats me how you can add significant functionality to a wristwatch-sized display.
Meanwhile Microsoft continues doing everything wrong and Samsung continues doing everything right.

2013 looks like being a particularly interesting year.



  1. No one will argue that 3D is cheaper to make than planar at the same geometry so, if 3D doesn’t yield, and planar does then Intel is well and truly f****ed, grax

  2. IBM announced 14nm FD-SOI in 2014; they will provide part of the answer not ST cause in Europe managers are not capable to take risk at all. (cultural barrier)
    Perhaps Otellini has been fired cause he has triggered this race to 14nm tech in order to compesate x86 CISC low perf on power respect to ARM.
    The move has costed a lot of money for the entire industry, pushing a lot of pressure on CAD industry as well without opening new markets … now it turns out that Intel 3D is not as stable, yield probably not so good as Samsung, Glofo and TSMC in 3D (see announces on Intel Ireland)
    If this turns to be true in next years, then Intel processors will be much more expensive and ARMv8 will get a big part of the server market meaning intel’s will loose good part of the stakes in this market becoming a sort of IBM
    David what’s your opinion?

  3. I wish I knew more about the Chinese players, Mark. Spreadtrum are exceptional, ZTE and Huawei seem to be running into trouble on the political front. I’d love to see some major semiconductor innovation coming out of China – with all those brains you’d expect something – but so far there’s been no one of the ilk of a Bob Noyce, or a Ted Hoff or a Ross Freeman, or a Bob Widlar. But then, mind you, there was no one of that ilk from Japan in its glory days. I do hope Asia is not going to be a disappointment

  4. No thoughts on the impact from Chinese players in 2013?

Leave a Reply

Your email address will not be published. Required fields are marked *