Electronics Weekly Electronics Design & Components Tech News
Spreadtrum gave ST-Ericsson a clobbering in the wireless market last year – one growing 95% the other sinking 28%.
The contrasting fortunes of Spreadtrum and ST-Ericsson are pointed up in IC Insights’ ratings for the 2011 fabless semiconductor industry.
‘China-based Spreadtrum was the star performer among the fabless IC producers last year, logging an amazing 95% sales jump,’ says IC Insights, ‘ Spreadtrum supplies baseband processors and highly integrated CMOS RF transceivers for wireless applications with over 90% of its sales to China-based customers. Spreadtrum has risen from being ranked as the 67th largest fabless IC supplier in 2009 to 17th in 2011, a jump of 50 spots in only two years!’
‘Spreadtrum and ST-Ericsson are both heavily involved in the cellphone marketplace,’ adds IC Insights, ‘in contrast to Spreadtrum, ST-Ericsson’s IC sales dropped 28% in 2011 after declining by 9% in 2010. The difficult times at the company led to a restructuring in March of 2012 and it is rumoured that this was a prelude to eventually putting the company up for sale.’
US-based companies took eight of the top ten places in the fabless rankings.
Combined, the top 25 fabless IC suppliers represented 80% of the $64.9bn in fabless IC sales last year, up 4 percentage points from 2010.
15 of the top 25 fabless IC companies outperformed the total 2011 fabless IC sales growth rate of 4%.
8 fabless companies registered double-digit increases in IC sales last year.
Last year was the first year on record that total fabless IC company sales growth did not outperform total IC market growth (a 27% increase for fabless IC companies versus a total IC industry growth rate of 33%).
This year, fabless IC suppliers were back on track in 2011 as worldwide fabless company IC sales increased 4% as compared to total IC market growth of 1%.
Overall, these top eight performers increased their combined sales by $4.5bn last year as compared to a $2.4bn increase for total fabless IC company sales in 2011.
It is obvious that these eight companies are now prime targets for the foundries, says IC Insights.
The Top Ten Fabless Companies in 2011.
Qualcomm 9.9
Broadcom 7.2
AMD 6.6
Nvidia 3.9
Marvell 3.4
MediaTek 3.0
Xilinx 2.3
Altera 2.1
LSI 2.0
Avago 1.3
Top Ten Fastest Growing Fabless Companies
Spreadtrum 95
Dialog 77
Qualcomm 38
MegaChips 35
LSI 26
MStar 15
Avago 13
Nvidia 10
HiSilicon 9
Semtech 9
Sudden movements are not expected, because, the b***s of most Swedish STE engineers are firmly in the hands of their banker, boss and b***h. Remember: We are the people who let the Germans operate our railways and ore mines during WW2. Business as usual for the cowar…, Swedes I mean.
Also, who in their right mind would like to give up on this employer renowned for providing execution excellence and deliverer of fantastic telecommunications products?
“he sites in Sweden are mainly working on the modem development, where a cut now would be very difficult – as the next generation modems need to roll out”
Have European chip designers lost their B***s?
I can’t imagine any other reason for the above quote, I mean dont the engineers that control their own fate?
If a market for the product exists and the present employer doesnt react quick enough than the opportunity passes onto the engineers, its that simple! Oh course the engineers need to have a pair otherwise all is lost.
Different sites has different specialities. The sites in Sweden are mainly working on the modem development, where a cut now would be very difficult – as the next generation modems need to roll out. Once they are out the door, it will be a different situation of course. It is all down to the management to decide priority what is strategic important for the company. Also, don’t forget that Sweden has been cut several times before!
Agreed – but we’re talking about what is really a contract design house here, not an international standard R&D house.
Here is the list of European ST-Ericsson sites:
it’s a combination of former NXP, ST & EMP sites
– Paris
– Rennes
– Le Mans
– Grenoble
– Sophia-Antipolis
– Bristol
– Basingstoke (closed)
– Daventry
– Dublin (closed)
– Zaventem
– Grimstad
– Rabat
– Zurich
– Casteletto
– Turku
– Tampere
– Nummela
– Linkoping
– Lund
– Nuremberg
– Nijmegen
– Eindhoven
Perhaps less if you possess a few patents of your own?
You pay the Qualcomm tax come what may – unless you go illicit like the backdoor chips from another well known manufacturer. I used to describe Qualcomm as a patent licensor with a small IC division tacked on but they have grown the latter well in recent years.
Given the EMP & STE track record of finding customers for their modem technology. Well, maybe it would be better to flog it off before its expiry date?
HTC seems going “vertical” in the supply chain, probably fed up with paying the profiteering qualcomm-tax?
Good points [Anonymous] it all depends on whether ST-E sees itself as a competitive company or a techno-charity. If the latter, then let the French look after the French; if the former, then keep your best engineers and lay of the ones who aren’t contributing much. It seems that ST-E management has made its choice about who and what they are.
I looked at that UNSA web page and it makes me crazy. Seems as they even comment presidential election in France. What the hell this organisation is doing? Do they want to help to make a business or simply substitute management of STE?
Actually it’s more like 1000 lay-offs and 700 reintegration inside ST.
http://www.unsa-st.com/st-ericsson-le-debut-du-demantellement,771
Belgique : Zaventem
Chine : Shangai
Finlande : Tampere
Suisse : Zurich
et Angleterre : Bristol
Europe massively hit 🙁
Yes Bitter, this could indeed be a cunning plan to hand the modem business to Taiwan. On the othe hand, the ST guys may feel thEy can make something of it under their exclusive non-Swedish management control, while I suppose it’s a strategically valuable capability for ST to keep. But no, I have no knowledge of what they’re thinking.
Please don’t forget Finns. In 2007 Nokia outsourced its BB and RF ASIC development teams to STM and now it seems that it’s their time to go… Even though only one site is to be closed this time it’s going to be a death kiss for the rest of them!
David, you bring us to an interesting discussion ?
Can you prevent a country to give subsidies under the condition that employment is preserved in that country ? Would you call it protectionism ? Not being French, I obviously don’t like it.
But what is the alternative ?
Assuming that an European law would be created “illegal to do lay-off programmes based upon on discrimination by nationality”.
Then, how should the lay-off be done (without going in the politics of how the lay-off should be spread)? Reduce the number of people in each individual site (leaving STE with many small site).
Isn’t the real problem of STE that they should have addressed this at the time of the merge, not 3 years post-mortem ?
Everyone seems pleased in Sweden and France after the positive news. Some socialist solidarity in action here I suppose?
There are some 400 consultants in Lund, I wonder what will happen to those guys and gals?
I also suspect employees not working with the modem specifics will be transferred to the ‘loving’ care of the French at STM? Someone correct me if I’m wrong.
David, do you got any insight regarding the rumored coorperation between HTC and STE? The modem business perhaps is going to the Taiwanese after the next round of layoffs?
So no French lay-offs, no Swedish lay-offs but 1,700 lay-offs, Bitter, sounds bad news for the Dutch and the English.I would have thought the Brussels bureaucrats would have fashioned some regulation making it illegal to have lay-off programmes based on discrimination by nationality.
And the Swedes behave like the Swedes…
So the farce seem to continue in Sweden, for the time being, as well. No layoffs in Lund according to the local news.
I’m in Silicon Valley at the moment and they say people are hiring like hell, [Anonymous] I assume the same is true in Southern California so I’d be brushing up contacts in Irvine and San Diego. Yes, the French are behaving like the French. The election is a great result. IMHO what Europe needs now is a solid dose of Socialism. The French are showing us the way – Vive Les Francais is what I say. The only problem is: will the anti-democratic EC elite accept the result of the French election (Assuming Hollande wins) or will the EC big-wigs order the French to hold the election again?
unfortunately it does not sound good for the Bristol people, and I have some doubts on the mid/long term for the Daventree people as well.
David, now is a good plan to come up with an alternative for these people !
And you are right… all closed sites are outside of France. Hmmm, any correlation with French elections ?
Well one of Maxwell’s equations was Faraday’s Law propounded 1831 and before Faraday there was the work of Romagnosi 1802 and Orsted, Robertl, so you could say 200 though it might have been better to say 180 years.
Where do get “200 years of European wireless pre-eminence” from? Maxwell’s equations are only 150 years old, and ‘Hertzian waves’ about 25 years younger!
My impression, Robert, though I may well be mistaken, is that Spreadtrum went out and asked potential customers what they wanted and delivered it. Other companies – Icera for instance – concentrated on the latest and greatest and then look around to find someone to buy it. My impression with ST-E is that it has a lot of research guys but not so many sales guys. But telecoms is a;ways hell – approvals and testing etc all takes so long.
Isnt it interesting to see how many of the comments relate to STE and how few involve Spreadtrum. I mean just how did this company do it? They were down and out 4 years ago, begging for a financial injection from rivals and yet today they are the fastest growing semiconductor company. I don’t know exactly how many engineers Spreadtrum has but I think it is fair to say it is probably less than 1/10 of STE’s workforce.
Oh well keep arguing about STE, I’m certain you’re all correct, but in the mean time don’t be surprised if high tech Investment companies shift their focus even more towards Asia.
Unfortunately for most of us, semiconductor design is ultimately just a ponzi scheme, we really profit from latching hold of the coat tails of a high flying company, rather than by receiving the wages necessary to make the venture compete against a more mundane government style jobs. The ponzi scheme only succeeds when the external observer sees something exciting that he wants to be a part of. So if you think about it in these terms, what kind of external investor would want to have anything to do with STE?
Yes Bitter let’s have a whip-round to send a couple of lbs of wacky baccy to Agrate
Maybe the ST&E corporate doctor should prescribe some Jobsian LSD “medicine” for the numb minds to get innovation going again? 🙂
Yes Bitter, and there’s that GB Shaw quote about all progress depending on the unreasonable man. Steve Jobs was suspicious of reason – said gut instinct was better. I hope these companies survive, but to do so they will need more than reason.
“Imagination is more important than knowledge”
– Albert Einstein
In a few years time I am sure we’ll be trolling David’s blog regarding the demise of STM and Ericsson.
Good point Bitter it’s like the ENA-effect in France – all the top people go there, so they all think alike, so when they get it wrong the effects are amplified. Same for Harvard Business School which is now admitting its past teachings have screwed up a lot of US industry. As you say, it’s the guys who see the future correctly and have the balls to pursue their vision of the future against the general consensus, who grow great businesses. I somehow doubt that anyone like that has survived the corporate culture at ST/ST-E.
Given the merits of the clientele managing STE I would argue that intelligence and education isn’t sufficient for running a successful business.
The success of Apple and Microsoft shows that school drop-outs can eclipse a highly educated, elitist and dullard corporate governance.
That’s reassuring, Adele. I would have expected a more up-front approach. ST-E is the last significantly-sized wireless operation in Europe – the heir to 200 years of European wireless pre-eminence – and it requires outstanding leadership to save it. But I’ve seen no evidence of leadership – interviews, actions, statements of intent etc. And he comes out of the risk-averse, cost-driven ST culture which sees ST now at the same revenue run-rate as in 2005 when Pistorio left. I really hope I’m surprised on Monday by an ambitious and imaginative plan to expand and grow ST-E.
Dave, not sure why you have it out for Lamouche. You say he’s a restructuring expert, but in fact his big restructuring success was Group Bull — which is perhaps more of a miracle than anything else. Other than that, he has a PhD in Semiconductor Technology. He started out in R&D at Philips. He was ops director for Motorola in the US for a while, then went to IBM as vp for ww semi ops. Clearly he’s a massively smart guy. To reduce him to “restructuring expert” is to miss the point that when it comes to technology, he clearly understands what he’s talking about.
I don’t think one can count open source as having a development model. Not saying it doesn’t work but companies need a model that produces a new, albeit derivative, sellable product every 12 months.
But the points you mention are the same ones I think are killing STE.
Linux would be a prime example of a huge and succesful distributed development model spanning a myriad of sites?
STE’s pitfall is the applied vertical bureaucratic brakes and with that a cross-site mudslinging, thereby making it impossible to cooperate effectively.
It simply cannot be efficient to design different parts of the same IC at so many different sites. Qualcomm has had to take on new sites as it’s expanded but they are almost all in San Diego.
No wonder STE can’t get anything out of the door !!
Please don’t ignore the STE WLAN team in Bristol.
>
You’re joking ? Only 1 ??
Let’s do some math…
#sites: 44 (from Q4-2011 investor call)
#products: 5 – 10? ( guess based upon # recent products / product families on the website, discounting ‘simple derivatives’)
=> #sites/product: 4.4 – 8.8
=> #sites/chipset or platform (assuming 2 chips per chipset-platform): 8.8 – 17.6
Of-course the products are pretty complex, explaining the large amount of engineers per product.
I believe we can easily say that 1 super-combo chip corresponds to the equivalent of at least 5 chips from 10 years ago.
I thought Bristol just did STBs and AV. Is there any STE work there ?
As for STE itself, one does wonder how many sites it needs to develop a chipset. One possibly ?
already 3 sites closed in UK/Ireland ..
-ex-NXP/Glonav site in Dublin
-ex-ST site in UK few years ago, people moved to Bristol
-ex-EMP site.. see David’s blog from last year
partially offseted by hirings in Singapore
You’re right Bitter, low cunning beats high thinking any day of the week.
Way too idealistic David! I would have believed that all your reporting and industry experience taught you otherwise?
I have a better idea! Offer an unrealistic plan back to profitability and relevance to the STM & Ericsson bean counting careerist clerks. Then place your buddies in the top spots using a ruthless ‘Ed’ shenanigan scheme.
Then watch your stagnated salary go nuts. It worked for the previous gang, it might work for yours?
Yes indeed [Anonymous] I shall be increasing numbers there substantially.
I’ll just have to reinstate the sacked ones Mike, the French exodus will leave a lot of vacancies.
sounds good for Bristol & Daventree
are there any Brits left ?
Well there wouldn’t be any more sackings of British engineers, Mike, but watch out the French!
Sounds like you are just the CEO they need !! As Alan Sugar says “you’re hired” 🙂
Yes, Anonymous, I can exclusively reveal that the April 23rd annoucnement will include: An announcement of the success of 28nm sampling out of TSMC which has led to widespread customer acceptance and substantial design wins at Nokia, Samsung, HTC and others; an accelerated programme to 20nm; a new product programme including a new chip-set for Windows on ARM PCs which is gaining traction in the market; new platforms which substantially beat the power/efficiency of Qualcomm/MediaTek,Spreadtrum chip-sets and a commitment by top management to take a 50% drop in salaries until they return the company to profitability. I wish I wish. In view of the fact that one of the top two guys is described as a ‘restructuring expert’ and the other is an accountant I regret to say I expect a French Surrender with sell-offs, dismemberment, lay-offs and an admission of defeat. I wish most intensely that I am wrong.
Can omniscient Manners throw some light on what’s in store for STE on 23rd April !!