Electronics Weekly Electronics Design & Components Tech News
In a year in which the semiconductor market is expected to grow 30%, NXP looks as if it will grow at zero per cent.
Q1 sales were $1.165 billion; Q2 sales were $1.2 billion; Q3 sales were $1.2 billion. NXP says it expects Q4 product revenue to be flat with Q3.
NXP’s biggest business unit – high performance mixed signal (HPMS) which represents $715 million of Q3 revenues or 70% of the business, was hit by capacity constraints.
‘Compared to the prior quarter HPMS revenue was essentially flat as capacity constraints limited growth in certain identification, microcontroller and automotive areas’, said NXP.
Why’s that? Well NXP cut capacity back massively last year and this – selling its fab in Caen, France in June 2009; closing its production facility in Fishkill, New York in July 2009; closing its Hamburg fab in January 2010; and starting to transfer process and products from its ICN5 and ICN6 facilities in Nijmegen, Holland which are scheduled to close in 2010 and 2011, respectively.
Now, says NXP, its fab capacity utilisation was 99% in Q3.
OK. NXP has improved its margins, but that’s not much consolation when you can’t make product.
‘Capacity expansion plans are underway and begin to impact later this year and throughout 2011,’ says NXP, ‘the largest investments are in our SSMC wafer fab in Singapore which primarily supports the high growth areas of our HPMS business.’
High growth?
Well it would have been if the axe hadn’t been wielded so enthusiastically.
@Mike
Allocation is a very complex subject and double ordering or early ordering will not necessarily change your allocation. You really need to understand how important you are to the Fab and who the fab is betting will win in your market segment. Unfortunately even this will not guarantee supply, the fab must want to win in your segment more than it wants to take market share in another segment. Not easy to know where you stand.
Regarding over 70% GPM products, they are still common for specialized SOC’s but the segmented blended target GPM is now usually 50%. This means that any company with over 15% R&D costs is in trouble, at over 25% you are in serious trouble.
David, I’ll give you a quick lesson in Mixed signal SOC chipset market analysis.
Dominant player has over 50% market share
Second player has 20% to 30% share
Third player gets 10% to 15% share
All other players split the remaining available market share.
there are very few exceptions to this rule and the TV chipset is certainly not one of the exceptions…
Well Robert I see from iSuppli that Trident has 13.8% market share in TV processors making it the No.1 supplier, and is also the No.1 supplier of LCD-TV controller ICs. So maybe the NXP link-up is delivering something, and the bull is giving Jove some competition..
We have been shouting out the wafer allocation point for years. The industry is painting itself into a corner where one company can decide if you live or die and inevitably there will be others who shout louder than you for allocation.
I would also like to add that I don’t expect there will ever be a 70% GPM product in this field again – it is too mature now.
Finally one of Trident’s problems is it is too top heavy and needs a clearout at the top level, not lower down.
In my opinion, Trident would have been on a quick trip to the chop shop if it had not merged with NXP. It had no viable products, no roadmap, big losses and loads of cash, plus several Vultures capitalists were actively running the ruler over it. In this sense the NXP BU debt and dilution saved Trident. NXP home BU, on the other hand had some new products and a reasonable roadmap, but KKR was sick of funding it’s ever mounting losses.
So the merger was one of necessity for both companies, KKR got a viable partner with plenty of cash to pay / fund the Home BU severance (make sure it did not recourse to KKR or Phillips) Trident bought a future.
Unfortunately the reason the Trident had no viable products was the same reason the NXP home BU was not profitable. A syndrome, known in some quarters as:
“****ed by the M&M’s”
If Big M is a vicious competitor, than Little_M is certainly it’s child.
Today, Trident desperately needs a breakthrough product with high GPM(over 70%). The problem is their competitors know this! They will make sure that every socket which Trident wins, will ultimately be profitless. Look very closely at the reasons why Trident lost the Sammy socket. I think you’ll find that the M&M’s boot prints all over the crime scene. (hint: understand how allocations work when wafer supply is tight)
From my observations, Silvia simply doesn’t “know the game” and doesn’t even realize when she is being played. A very dangerous combination at any time, I fear it’ll be a deadly combination with such adept competitors.
Maybe Silvia needs some Latin lessons, I’d suggest she start with:
“Quad licet Jovi, non licet bovi”
There is a fascinating document on the history of the Southampton site at http://epepnet.co.uk/History.pdf.
In the 1980’s manufacturing went to the far east. People said this was not important as the UK could concentrate on its higher valued design skills. Now in 2010 design has gone too.
It sounds like a shockingly cynical manoeuvre to avoid redundancy costs, Clanghead, KKR’s ownership record at NXP is appalling. I hope the current EU legislation about private equity activity in Europe will prevent this sort of thing happening in the future.
It was announced last week that Trident Microsystems at Southampton (ex-NXP Business Unit Home) is to be reduced from 31-Dec to 3 (three, count ’em) engineers. They’ll be working “from home” and the site will be shut.
When NXP took over Philips Semiconductors in October 2006 there were about 550 people working there.
NXP cock-up in cutting capacity too much? Shurely shome mishtake!
Well Yes Papa, but the market was so bad in Q309 and so good in Q3 2010 that it would have taken a very determined effort by management not to have y-o-y growth.
My point was that to have flat sales in a 30% market growth year – and blame it on lack of capacity – after fabs have been closed in France, Holland New York and elsewhere suggests that someone made a cock-up in cutting capacity by too much
Dear David
NXP announced flat sales quarter to quarter, but +25% year to year.
At least they helped Trident Microsystems, the company NXP gave thier TV and STB SoC businesses.
From the transcript of Trident’s last quarterly results call it looks like they lost the business with their own TV SoC in Samsung and they now rely on sales from PNX85500 and PNX5120/30 TV SoCs plus STB SoCs all acquired from NXP. But this was enough for them to break-even.
Is there anything left of the old NXP SoC team in Southampton?