It could also be that ST-Ericsson is not now in suitable shape to be
sold and requires a new plan and more cuts by CEO Lamouche. ST-Ericsson
is still trying to present itself as mobile processor vendor, which
requires a lot of software and motherboard work on top of the merging of
modem and application processor. In a market dominated by a few
operating systems – iOS and Android – it would seem that ST-Ericsson's
greatest value lies in its LTE modem capability, which is only a portion
of its current operation.
We asked back in July 2011 how long
ST-Ericsson could carry one, but we never dreamed it would be two years.
Earlier this year, we suggested a sale to a well-heeled Chinese company
might be one of the quickest and easiest options. Things are of course
made more complicated by having to satisfy the requirements of the other
anguished parent, Ericsson.
The irony is that FD-SOI, a
manufacturing process technology that was more or less engineered by ST
to help ST-Ericsson compete in the mobile device market remains a
potential key to the rejuvenation of STMicroelectronics. There are signs
that FD-SOI could be a real winner against bulk CMOS and FinFET CMOS
processes at 28 and 20 nm, especially in terms of low power consumption.
the mobile processor market has been all but won by Apple, Qualcomm,
Samsung and a few Asian competitors. ST is more or less out of leading
edge CMOS manufacturing. At least FD-SOI gives the chip maker a viable option for the future.
Good point Peter, wanting to sell and being able to sell are different things...but in business, real estate or life in general at the end it boils down just to money...how much are they asking for that money losing business?
Europeans, unfortunately, cannot business decision in a decisive and convincing manner. This behaviour appears to be very consistent over the past many years, hence STM / Infineon / NXP keep becoming smaller, smaller, .... If nothing drastic happens in the opposite way, STM could eventually goes the way of Renesas, i.e. dinosaur. In many ways, they are very similar, both a combination of many companies.
Once ST-E is left on its own to fight for survival, it may go to Samsung / TSMC / SMIC or anyone who can offer cheaper foundry services than the 28nm FDSOI. In particular, what is STM's roadmap for FDSOI beyond 28nm? AP with integrated multi-mode modem cannot practically survive for long on 28nm! Given that STM is NOT #1 in microcontroller market share, it won't be able to afford to keep the 28nm FDSOI especially if ST-E goes for other foundries because of cost reason (though a distinct possibility at the moment, one can never tell when and if it happens).
It appears that, at the end, semiconductor companies from US, Korea, and China/Taiwan could be the only ones left standing.
So why did (BST) he tell announce a deadline? Maybe to secure thousand's people job for a while not having yet an alternative. In such deep crisis moment we should sometime start thinking also about people rather than just financials.
I don't disagree that financials should be tempered with an awareness of social impact. But no business can fly in the face of the financial imperatives over the long term. The risk is that moving slowly may preserve some jobs in the short-term but put many more jobs in jeopardy in the longer term.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.