LONDON – Something has changed in the psyche of semiconductor manufacturers according to Malcolm Penn, founder and principal analyst with Future Horizons. They are no longer building wafer fabs in anticipation of demand. "Forget fab-lite, welcome to the fab-tight era," said Penn speaking at a one-day seminar on his market forecast for 2011.
Penn presented a 6 percent growth figure for the chip market in 2011 but he feels that the fundamental situation has changed. That change will drive good news for the foundries - and not such good news for everyone further down the supply chain.
In the past chip makers and pure-play foundries have tended to build wafers fabs in an attempt to be one of the first with significant manufacturing capacity at new manufacturing nodes. This has led to the risk of oversupply to the market and boom-bust cyclicality. But in the latest cycle, overlaid with the general economic crisis of 2008-2009, capital spending on semiconductor equipment has been conservative, Penn said. Thismust lead to an undersupply situation, increasing average selling prices and problems for chip companies that do not have control of their own manufacturing, he said.
"We've been running [manufacturing capacity] maxed-out for three years. We are now running in a fab-tight mode," said Penn "The chip makers are building fabs after the demand, and this changes the rules. The industry just hasn't realized yet."
Penn also described the coming years as "pay-back time" as he expects leading foundries such as Taiwan Semiconductor Manufacturing Co. Ltd. (TSMC) to use their strong positions to raise the cost of wafers. Some observers believe previously pure-play foundries may even be tempted to enter the market place and sell direct to leading OEMs.
Penn has been a long-time opponent of the so-called 'fab-lite' policy whereby established chipmakers have saved on capital expenditure by outsourcing some part of their chip manufacturing to foundries. This has usually been the leading-edge digital manufacturing leaving the chip companies to fill their older wafer fabs with less aggressive digital designs or analog, mixed-signal circuits. Penn has consistently described fab-lite as unsustainable in the long term and as just a way of describing a transition to fabless status.
This is not surprising at all to me. The cost of a new fab is more than most companies can bear and it does not make sense for the few that can afford it. Moving your wafer production to foundaries just makes sense. The problem is going to be, who gets priority when the fabs are full? The biggest customers will. The small startups and fabless companies will be at the mercy of the foundaries. But then, they always were.
I agree with dakeste that the cost of the fab is going to be so much that single company will not be able to afford the cost. But i think that this is just one scenario. Another scenario can be a monopoly (single or consortium) so that to be able to control the market prices better.
Agree. I think the few companies who are still holding onto and expanding their fabs while working on innovative products will win over the new and up coming players due to this scenario in the long run. It is neccessary for a more ambitious semiconductor company to retain considerable independance on the manufacturing front.
I disagree with the hypothesis that fab-lite is just a transition to fabless. Companies that make a variety of products that use different technologies have diverse fab needs. The analog mixed-signal group, for example, is always going to be using larger geometries than the group that makes processors & SoCs.
As a business trying to maximize your return on capital, which fabs do you invest in? The ones that build your mixed-signal products? The ones that require the latest process node? The ones that build the products with the most sales? Maybe instead you invest in those that build products that cannot be easily moved to a foundry, and leverage the foundries to handle upsides in demand for those products that the foundry is able to build for you.
Just because you have a diverse supply chain and use a mix of internal and external fabs does not mean you are in a transition to fabless.
It is all about economy. Each company will have to honestly calculate their cost of wafer price from internal fab vs. external foundries. Product lines that can get cheaper from self-operated in-house fab line should continue to do so. All others should seek a more economical wafer source which is foundry.
Key word here is "honest cost calculation" which needs to include good and bad times when the internal fab is under utilized.
If Penn is right, such a fab-tight era would be a boon for new foundries eager for market share...
Judging from their capex plans, i bet Samsung and GlobalFoundries are doing their best to supply customers who can't get their sweets from TSMC...
But not many companies have the process R&D to get into the foundry market at the leading edge. As you highlight besides TSMC there is Samsung, GlobalFoundries, then there is Intel.
The only other companies i see able to put down a shell are Toshiba, Hynix and maybe Micron who are working on the memory front.
And that's it. Seven 'active' large-scale chip companies. The rest are milking the fabs and processes they have already developed.
with 'new foundries eager for market share' i actually meant only Samsung and GF.
I agree that apart from these two there probably isn't anyone else who has both the money, the technology and the will to become a major foundry player.
I agree with Peter, at best there are maximum 7 foundries that can afford a new fab. With so much risk and such a high cost why enter that market? I bet that 7 will become 5 or maybe 3 in 10 years out. Look who builds airplanes: Boeing and Airbus. The rest is just small planes, niche markets. Semiconductor processing will be the same eventually...Kris
"Fab-tight" is great wording. SEMI's World Fab Forecast talked about exactly that end of November. See article: "Fab Capacity Back in the Black - uncertain outlook for New Fab Construction" http://www.semi.org/en/MarketInfo/ctr_042104
Some older or more specialized technologies perhaps. But in essence wafer production is about economies of scale and whoever makes on the biggest wafers in the biggest volumes has the per-chip price advantage.
The problem is you have to commit to spend $5 to $10 billion to do that. So it comes down to the cost of capital. In Taiwan TSMC was encouraged and supported has now achieved a leading position. In Abu Dhabi they are providing about $10 billion to give GlobalFoundries a chance of getting established.
That is helping to produce the GlobalFoundries' New York wafer fab but apart from that i don't see much reason for optimism about chip manufacturing "coming home."
OK....I can see production on 450-mm diameter wafers coming from say, Intel, Samsung and TSMC, for example. But only after a period of stand-off when each waits for one of the others to commit the funds and pay the price of going first.
So Floating Gate when do you see 450-mm production starting... 2015, 2020 or 2025 or later?....and a new player?....maybe in a niche ....but not in high volume, not on 450-mm wafers, surely?
450mm will not be possible until Intel builds D1X or TSMC, GF or Samsung builds a 450mm ready fab to house the tools. Even then where are the tools? Nobody, that I am aware of, has even built a prototype 450mm production tool. Scaling a production process from 300mm to 450mm is going to be very challenging, to say the least.
If we go by Intel's model, it will take them 2 years to get 450mm process up and running once the tools are in place. Maybe 18months if they can transfer the previous process from 300-450 smoothly. Regardless that time is after the tools are developed and manufactured en mass.
One thing that I don't like about this fab light, tight or whatever strategy, is that we do not have the R&D capacity for new technologies to be developed and incorporated into new devices. Foundries are all about using the same few processes to fab logic, dram, flash, bulk ICs, etc, and they are only looking to develop technologies that will fit those needs. What innovative technologies are we missing out on because companies cannot invest in technologies that are a little more out there? I mean things like superconducting quantum computation or other interesting analog technologies that come from cutting edge device research in academia. These types of technology really need an industrial approach to get them working, but the prevalence of the fab-light-tight model will have them labeled as too risky and unfundable.
Well, fab-light or fab-tight or fab-less, right now there is more concern over where the worlds economies are heading in the short term and that leads to a lack of investment until there appears to be more short term certainty.
I believe there is a healthy market for MEMS devices which do not need a state-of-the-art fab and the associated big price tag. These along with 3D chip integration via stacking are expected to provide economies of value as an alternative to few IDM's that afford economies of scale.
The 3D integration though may be more a packaging and assembly play at the backend than the fab at the front. How ever, the 3D 'flows' are still evolving, it will be interesting what 2011 brings.
Dr. MP Divakar
Many chip company has gone to fab lite because of economic downturn and advanteges of foundry services. However as soon as the economic situation comes back normal i dont think theyd be gone for fabless phase. The nicest example of fabless company model is AMD, where they were battling against intel. No need to talk about who is the winner of the race (ok let say AMD had some other issues but main thing was lack of production capacity/advanced pricing)
Yeah, its true that building up a new fab gets more and more expensive and few can bear it but the rule is the same. Who is really eagered to be involved and win the game. Abu dhabi has money and would like to power up anybody who would be in their side. Who knows about Chinese fellows and their dreams
I still dont think tight capacity and foundry fact would effect future of semiconductor markets that dramatically, its rather effecting speed of release of innovative electronics and of course the increasing the cost/value.
in any case who has the money and will, can still gear up for higher capacity if their need increases. Example? TI and their vertical implementation of fabs, toshiba and now sony...So who has the money still can get fab capacity without need to deal with foundry.
Join our online Radio Show on Friday 11th July starting at 2:00pm Eastern, when EETimes editor of all things fun and interesting, Max Maxfield, and embedded systems expert, Jack Ganssle, will debate as to just what is, and is not, and embedded system.