Apple should build a fab

8/26/2010 7:57 PM EDT

I am not saying Apple should build a fab today. But what about in the near future? I'm sure Apple could find the expertise and money. Thoughts?

8/26/2010 9:09 PM EDT

Personally, I hold the opinion that building a fab because of Samsung's rise as a competitor is not justifiable. Apple could always go with other foundry players who are not in the path of conflicts for their ARM processors. I've always seen Apple as a design company, something like maintaining a competitive fab should really be left to the professionals.

8/26/2010 9:24 PM EDT

Chandler, I agree. Getting on the semiconductor manufacturing train and following Moore's Law to the end of the line is way outside Apple's business model and it's expertise. If anything were to happen to the Apple-Samsung relationship, there are other foundries...each of which would jump through hoops to win the Apple business.

8/27/2010 2:32 AM EDT

True. But as I stated, Apple would get no preferential treatment or special pricing at TSMC or GlobalFoundries. Apple would just be another customer even if TSMC pulled out the red carpet. That would bug Steve Jobs.

8/27/2010 2:41 AM EDT

I am for the idea that Apple should have some controll of its chip supply that has the longgest time to bring up an alternate. However, I believe there are many other options tha 'building' their own fab, how about leasing some capacity from those 2nd tier foundries (SMIC, UMC, GF..) that are dieing to get some popular products to fill their empty 300mm fabs (if not empty now,they will), with this, they can also leverage Samsung chip price (not that Apple need this now given their product popularity and margins).

8/27/2010 2:44 AM EDT

Apple, now the darling if Wall Street, would be hated by financial markets and investors if they took on the huge capital expenses and risks of a fab.

8/27/2010 3:13 AM EDT

To a large degree Apples success is a result of its "closed" system approach. HW and SW are designed together and only Apple gets to use Apple technology. This model has been very successful bot financially and customer wise. However, Apple is light years away from pushing the number of devices that would be needed to justify investing in its own fab. To run a successful fab Apple would have to offer service to fabless companies where many would likely be in similar market segments .. and I do not see that happening anytime soon .. if not ever.

Value lies in design, not technology

8/27/2010 6:31 AM EDT

It might help Apple to be tension free and will help the supply chain to make sure products are always available in the market. May be apple should build a fab if they can't innovate and bring a new product into market which could succeed like ipod/iphone. I believe all the Apple customers(say it as apple crazy) are not really worried if there is a delay in buying a gizmo. Success of Apple products is in their innovation and faster time to market for the first time.

8/27/2010 6:48 AM EDT

As Apple does the chip design and orders substantial amounts, getting a second source at favorable conditions should be no real issue. Therefore, Apple would have no problems remaining fabless. At the risk of detouring the thread, my concern, was I Mr. Apple, would be displays. Here second sources are difficult to see.

8/27/2010 9:20 AM EDT

A lot of Apple's chip suppliers (Broadcom for example) don't even have fabs. It's not just about money, it's about volume necessary to ride the Moore's Law wave. Even iStuff doesn't have that kind of volume.

We're really in an age where fabs are consolidating due to high costs and low margins. Unless there's some very very clear reason (vast improvement in performance, power or area), it's just far too easy and less costly to switch foundries than to build your own fab.

8/27/2010 9:25 AM EDT

Now Apple starting to experiment with different materials and making plants based on that, possibly. Areas such as MEMS devices, screen technology (they can't be very happy with the power-hungry, sunlight-washed-out LCD) or an investment into graphene based semi-conductors I could see. Something that would seriously give their products an advantage but doesn't require volume since nobody else is doing it.

8/27/2010 11:23 AM EDT

Apple may go for fab if they have path breaking innovative ideas for how ideal fab should be and if they remain first or second in that field in terms of technology. Along with making individual parts like A4, they may be conceiving 3D packing of parts (at PWB level and at IC level). This helps them considerably for providing complex functionality in very limited space. They may consider monolithic multi chip module or SoC which has many or most parts into one big IC package. They may need this to implement their chimerical concept and maintain major lead over competition.

8/27/2010 11:48 AM EDT

This idea is a sure bet for failure. Apple does not get any competitve edge building a fab for themselves. Apple's success hinged on continuously bringing out new product concepts and innovations based on the pradigm shifts in which ways we humans use the products. Also having a company like Samsung continues to support Apple's supply chain is a synergistic. Apple can tab into Samsung's R&D, product design and get a time advantage before Samsung introduces similar products - which is also a good thing bacause the market size for Apple-like products expands considerably and this creates additional demand for real Apple products. Apple can't cover the world alone..

8/27/2010 12:27 PM EDT

Mark, I agree with you to some extent! There is a good case for Apple to have its own fab. How ever, it does not have to be a next generation fab (sub-22nm node) where Samsung already has a significant lead in expertise and experience. It will take years for Apple to catchup with the leaders in smaller technology nodes while the cost of building the required fab keeps going up.

Another option would be use to the current technology (28 to 22nm) to build innovative products through vertical integration. As "@hm" also says, 3D stacking of different chip functions brings good benefits to handsets and smart devices where real estate available is always challenging. Apple can really make a distinction here in its products. Logic, memory and sensory functions integrations in 3D stacking makes a lot of sense for Apple to keep its lead in technology.

Unfortunately I have not seen much participation in 3D chip development efforts from Apple (or they may be closely guarding it from public access) whereas Samsung and Qualcomm have released many studies. There were several 3D-focused events at the last DAC in Anaheim where Apple was non-existent.

8/27/2010 1:40 PM EDT

I just think Apple has to own its team of chip designer but not a fab. Building a processor doesn't need to own a fab as the manufacturing of such fine line CMOS process is basically standard to almost all foundries. The only important thing is to have the control on the architecture IP. If Apple is good in developing the A4 processor with its own IP, they can put it to anywhere for fabrication and get the lowest cost possible.

8/27/2010 1:48 PM EDT

Agree with all comments. But to me, Apple wants control of its own destiny. If or when the world goes to TSV-based 3-D chips--perhaps at the 16-nm node--Apple may want a fab then.

8/27/2010 3:22 PM EDT

Along with technological aspects, Apple may like to do this to control and lower overall price of BOM significantly. Most of savings will be passed to consumer to keep Apple products competitive and also to bring to serve the middle and bottom level of market in developed countries. Real success of Apple products is when they can effectively make developing and under-developed country part of their market.

8/27/2010 5:14 PM EDT

Moment Apple is competing based on silicon, Apple will go down the history.

8/27/2010 6:32 PM EDT

How would you suggest that Apple start from scratch to come up with a process to make 32nm or 28nm or even 22nm and not infringe on patents from either IBM's Fab Club, Intel or TSMC to name a few. Would the licensing fees negate the advantage of having your own fab?

8/27/2010 7:53 PM EDT

Utter nonsense. If you understood Apple's business in particular and the dynamics of the silicon industry in general, you wouldn't even think about suggesting something like this.

8/27/2010 8:23 PM EDT

There are other ways to do this. A joint venture fab with a foundry comes to mind. In this scenario Apple is guaranteed a certain run-rate of capacity, but when it is not needed the foundry is free to sell it to other customers. It costs money but not as much as owning your own fab and having to develop or buy the process to put in it

8/27/2010 8:37 PM EDT

Too risky. Apple has done very well competing on design and high value branding. If Samsung adopts a similar model (unlikely) Apple could switch to another partner relatively easily, and there would be no shortage of suitors. If it ain't broke, don't fix it.

8/27/2010 11:16 PM EDT

Fabs are built over cheap labor and deep pockets for capital equipment. A fab can't just churn out A4 processors unless 1/2 the universe is going to buy them... There's got to be a much better reason than profitability for getting into the fab business such as divesting from petroleum or bringing nano-technology to the Arab World. Yes, an ego is a good reason too. But even a hubris is controlled by his board...

8/28/2010 12:03 AM EDT

I don't think Apple would build its own wafer fab.
It's too risky to invest a tens of thousands M$ ?
Instead , they would use a wafer foundry like TSMC, UMC etc..

8/28/2010 4:32 AM EDT

Apparently, 25 years of learning and re-affirming that being a system house and a sporting a vertical manufacturing system at the same time does not go well together. Let me repeat some of the main reasons:
1. Alignment of product to the process, and as a result: (a) rapid process development to stay at the leading edge; (b) Fluctuation in the utilization of a single market-segment demand, unpredictable and deadly.
2. Huge cash outlays every 1-2 years with a 5 year or longer ROI - major disconnect between the dynamics of the system and the manufacturing enterprise.
3. Management bandwidth. After all, if Steve does not or cannot fully understand the fab, both sides of Apple will suffer. You cannot invest billions and then ignore it.
4. The fallacy that assuring supply means owning manufacturing means. Long disproved as a concept.
5. Supposedly better prices - wrong! Is that why Intel goes to TSMC?

In short - not just bad, but really, really bad idea.

8/28/2010 1:31 PM EDT

Apple own fab? Steve Jobs is making a terrible mistake, if he ever tries.

As some had earlier mentioned, Apple gets a better deal by throwing contract for its A4 to any foundary. In this way, Apple only needs to manage sales, marketing and CE development.

Fab is a liability, unless you got volume for a long term.
How many giants can substain long-term profit with fabs these days?

Intel even selectively selling away its fabs this decade.
STMicro and Freescale as well.
Many Japanese corps had also done so.

You need a large volume to keep it profitable.
Samsung has volume.
Intel has volume.
AMD has volume.
STMicro has volume.
TI still kept some own fabs.

You can simply count how many giants still hold on their fabs for new and strategic products.

For matured and low-volume parts, they outsourced to TMSC, UMC etc.

FYI - a fab needs at least a decent US$500 million to build. That's not a small amount.

8/28/2010 8:54 PM EDT

I have to agree with the majority of the comments that this would be a bad idea. I'm not an expert on Fab houses, but what I've seen in companies constantly retofiting or selling off fabs as we shink the chips sizes, makes me believe Apple would be out the fab business in 3-5 years with major losses for what is really no business advantage. There really are no business reasons for this, unless Apple want to be a general chip manufacturer, which goes against everything Apple stands for.


The only other reason mentioned is ego. Dear God, if Steve Jobs will spend 4-10 billion dollars because of ego, the board of directors should vote him out now. I have to believe with his success, he would see benefit of buy vs. make that all of us see.

8/28/2010 10:26 PM EDT

The idea has some merits and yet is so far outside of Apple's management expertise as to possibly be ludicrous ( evne if they bought hired guns so to speak ).

My counterpoint to real men have fabs
( ie TJ's profitless bluster )

http://www.siliconvalleywatcher.com/mt/archives/2005/10/an_engineer_res.php

Still true today. STICK TO YOUR KNITTING, esp when the cpus in handsets are basically commodities.

And I am a process engineer, with silicon in my blood.

mark-nano.blogspot.com

8/29/2010 2:49 AM EDT

I agree with Jan.Mikkelsen that displays should be much more of a concern than building fabs. For one providing added products for the whole ecosystem of the consumer space is the only way to continue growing. There will be a need to get inside the display electronics to provide newer features and self actualization with mass appeal. I am not saying they will want to build the panels, but they will probably start designing and selling video TV products, integrated PVRs,even game consoles and so on. This alone will require major R&D. Samsung already has TVs with Apps and Apple needs to continue to innovate...I absolutely do not think Apple wants to get in the fab business, they could though multi source more aggressively to get supply stability.

8/29/2010 10:10 AM EDT

Samsung is a unique player in the semiconductor-OEM ecosystem. There are always two sides to the story: How much business you give to Samsung and how much you take away from their markets. I guess, in case of Apple, Samsung would like to sustain the relationships and not pull off any adventure. Besides, entering the fab business would be too risky for Apple. Just moving on to a new foundry partner would be much more profitable.

8/29/2010 11:12 AM EDT

To go to 32-nm and beyond for the A4 (and the follow-ons), Apple would need a high-k/metal-gate technology. Apple has four options. Option A: It could work with IBM's ''fab club.'' IBM claims to have high-k, which is still not in production despite a lot of hype. (Why is that?) Samsung is in IBM's ''fab club.'' So Apple could join the fab club and build a fab. Samsung would be a second source. Option B: Apple works with the rival camp, TSMC, which is also late to the game with high-k. Option C: Apple teams up with Intel. But Intel would not give Apple its high-k technology. Option D: Apples devises high-k on its own. If it can do that, who needs IBM, Intel or TSMC?

8/29/2010 11:28 AM EDT

I think the main point is Apple's products are not all 28 nm at this time. You have some 45 nm, some 65 nm, some 3x nm, some MEMS, some displays, etc. You can't control supply of everything with just a leading edge fab.

8/29/2010 2:03 PM EDT

A lot of interesting comments. We should keep in mind that Samsung has similar relationships with a number of other major manufacturers where they already compete head to head with products. Also, running a fab means you need to have the volume to make the fab economical so Apple would really need to become an IC supplier if they were to go that route. I seriously doubt if that is in the offing anywhere in the near future.

8/29/2010 4:26 PM EDT

As has been pointed out, setting up their own fab makes little sense. On the other hand, if they could gain an advantage by acquiring a company which had fab facilities and the requisite expertise they might. That's how they gained their A4 CPU design.

Larry M.

8/29/2010 4:32 PM EDT

I have seen companies struggling just to maintain fab. Especially when new technologies nodes are getting very expensive. It very tough be another Intel with successful in-house fab. IBM has good fab but got out of services and TI stick to design and got out of fab.

The investment on 28nm or 22nm runs in billions and it not so easy to justify the cost even if they save $5 per chip for 1 billion chips per node (life of one node is about 3 years so assuming they are going to sell 1 billion chips in 3 years).

For Intel gains from fab because they are one node ahead of market and for that they invested a lot in RnD.

8/29/2010 6:49 PM EDT

Apple has a winning model, turning a dying firm into a cash cow. Why should they change that model. Building fab does not work well for most firms. Ask AMD. I think Apple should focus on the knowledge part and lesser on the fab business.

8/29/2010 7:02 PM EDT

The only companies that should build fabs are companies that are in the fab business i.e one fab with many customers. To build and staff a new fab would cost Apple several billion dollars - every year if they want to keep up with the technology shrinks. If there is any downturn in business, Apple would have an empty fab and have to lay off a bunch of employees.
Apple is very good at negotiating contracts. They can go to any existing commercial fab in the world (TSMC, UMC etc.) and get preferential treatment. That said they may not want to go 100% with Samsung, but there are at least 4 to 5 other fabs they could use where there would not even be a hint of future competition.
Even Intel and other companies with fabs will use outside fabs for over capacity.

Patrick

8/29/2010 8:45 PM EDT

The comment about Texas picking up some of the tab for a new fab is overstated. The Texas Enterprise Fund was given $260M for the current biennium (2009-2010). So any funding from Texas would be at best a few percent of the cost of a new fab. Local infrastructure, local suppliers, and local workforce would be much more important factors in a siting decision.

8/30/2010 12:58 AM EDT

Why would it make sense for Apple to have the costs of their own fab? It would not. Most semiconductor design companies are going fabless these days because of the expense. The short and long-term costs would outweigh any benefits received from having a captive and secure facility.

8/30/2010 6:15 AM EDT

In today's world system integration and Fabrication are two separate complex puzzles.
It is more or less comparing "Apple and Orange" in present day context.

In apple gets into Fab then then definitely I would ask why apple has started to make oranges? ;-)

8/30/2010 10:53 AM EDT

I don’t think it’s necessary for Apple to build it’s own Fab. A striking example is AMD. Once AMD sold their fab business to GlobalFoundries, they are able to concentrate on what they do best, making processors. I think the new leaner and meaner AMD with it’s new lineup of BullDozer processor architecture can make a dent in Intel’s business.
So I hope Apple remains lean and outsource its Fab operations to Samsung :)

8/30/2010 12:12 PM EDT

I don't see with 8.8M iPhones shipped in Q1 2010 how Apple could possibly justify building a fab.

http://www.zdnet.com/blog/btl/apple-iphone-smartphone-market-share-surges-rim-slips/34181

Even assuming they had the money to do it against all good common sense, they wouldn't have the organisational competence to do it and it would certainly not add any value to the Apple brand.

Even a JV at the volume levels they are shipping, even if they outstripped Nokia in the smartphone segment at 20.8M units in Q1 wouldn't make commercial sense.

In fact a JV would deprive a cash-rich A-list customer like Apple of the ability to shop around when there is excess fab capacity.

8/30/2010 2:16 PM EDT

Business has changed in the last 20 years. The trend that I have observed is business is trying to focus on what they do better, letting the others to do what they are good at. I believe this is one of the reason there are more fabless chipmakers in the last 10 years. There is no doubt that some chipmakers, like Intel, need a fab since their success is highly relied on fabication technology. For Apple to build a fab for the quantity is still a question. I believe for the foreseeable future, Apple will be better off if they continue focusing on their innovation in the creating new markets.

8/30/2010 2:24 PM EDT

Apple has built very successful business model that doesn't rely on owning a fab. It is not clear to me why would they change that. As brought up in a discussion chain owning and opeating a fab is capital intensive biseness that is hard to manage. There are only very companies in the world that can do that effectively and we have seen number of players in this space decreasing annually. Som epredict there will be 3-5 fabs left in the future, TSMC, Samsung, Global Foundries, Intel...Kris

8/30/2010 8:49 PM EDT

OK. So Apple should not build a fab and leave their production future to Samsung (or TSMC or whomever). Is that such a good plan?

8/30/2010 9:59 PM EDT

In general, now is not a good time to build a leading edge (22 nm or 16 nm) fab. Especially with people watching technologies like EUV or TSVs, maybe Apple is watching too. You don't want to build a fab with technology that is planned to be replaced soon.

8/31/2010 10:58 AM EDT

With Apple's market cap of 220 billion they could easily buy Intel, AMD, all the leading EDA and IP vendors and all the major foundries and still have change left over. Sure it would have anti-competitive repercussions and would be a nightmare to integrate those pieces, but this just gives you an idea of just how insanely high their current valuation is.

8/31/2010 4:47 PM EDT

it all depends on industry trends. Apple has made a big impact n design n high volume brands. owning fab takes big chunk of investment that i do not think Apple is looking for now...who knows future industry trends make them think to own a fab.

8/31/2010 4:54 PM EDT

It absolutely make no sense for Apple to buy a fab house. The key innovation that Apple has is its design-ability and marketing.

People have put too much into its engineering which of all thing considered a PC in terms of electronic hardware is much more advanced that anything Apple has.

So what I'd like to say is it make no sense for an internal fab. Look at it this way, say Samsung relationship gets out of hand, ok, lets go to TSMC, or Charter, or whomever semiconductor and SS can go pound sand!

8/31/2010 7:04 PM EDT

Apple's Exec Mgt. Team has simply created some of this decades most creatively functional smart devices! A continuation of this development trend to other areas like smart home entertainment, smart home energy system; may elicit executive discussion of such a major capital expense like it's own FAB.
Apple's near-term capital spending -- manufacture more product in America! Exporting changes the game here and abroad! That would be as exciting and those smart devices they market!

8/31/2010 8:31 PM EDT

They said Apple should not get into the cell phone business. Tablets too. Apple is a risk taker. Fabs might be a different story. But it's a piece of the puzzle. Ask Intel, TI and Samsung. Then ask Qualcomm. Expect a surprising answer from them

9/1/2010 1:10 PM EDT

Agree it is a piece of the puzzle, but getting into cell phones or tablets did not require Apple to become capital intensive with captive manufacturing. Design and product development already were things they were good at and they could beef up the teams with needed expertise for a resonable dollar, all of which could be cut later if it didn't work out. Getting into the high investment of a fab is rather different, not to mention a different operational business.

Seems like fabless has worked out for Qualcomm, Broadcomm and others. Don't see how wall street would reward them either if they decided to build a fab.

9/1/2010 5:00 PM EDT

Apple is employing A4 processor in newly announced Apple TV. They may use them in latest iPods. After processor, Apple may have new idea for memory – memristor or similar and other critical ICs. If we recalculate their requirement with all these new applications, Apple may get enticed to have its own fab. Apple may have different approach to source, build at lower cost and importantly utilize it in more dramatically different ways. It will be interesting to see their novel ways in fab business.

9/2/2010 9:32 PM EDT

I think it’s a risky proposition. I’m not sure if building a fab is the best way forward for a company that’s better off focusing on the next big consumer product.

Apple and Samsung may no longer be the best of allies but that doesn’t mean that the former should jump on to the fab bandwagon. Apple may be better off inking partnerships with other foundry players in order to have a Plan-B in place if the Samsung deal goes bust.

IMO, Apple is better off as a fabless company focusing on the end product rather then get into the nitty-gritty of fab world. I agree with Siliconman that it’s a risk-laden prospect which can backfire badly for Jobs & company. If Apple wants to control its destiny, does it really need a fab? I’m not so sure.

- Keith Schaub

9/2/2010 10:57 PM EDT

People who think Apple should build their own fab have been drinking too much happy juice.

9/3/2010 5:11 AM EDT

If Apple wants control of its destiny, it needs control over the technology used in its products. This means they should invest more in their R and D efforts to develop technologies required, which may be cheaper than that offered by a third party. Ofcourse this isn't easy if there are third pary patents/licenses involved, but it will help reduce the cost of produc development. Running a fab is expensive and for it to be economical, Apple will need to sell a lot of their products to keep the fabs busy. Maybe in a few years when Apple have greater domination in the market, but not now I think.

9/6/2010 1:49 AM EDT

Hi Steve (Jobs)! I'm sure you've noticed this: Your chip/foundry partner, Samsung, has rolled out its tablet PC. It's an Android-based system. Samsung plans to sell 1 million units this year and 10 million next year, according to HSBC. Vodaphone will sell the product, which can make calls. Did you also see this? Samsung has increased its smartphone shipments to 25 million units, up 39 percent from its previous forecast. Last but not least, Samsung rolled out a notebook PC line. Steve, time to call TSMC and GlobalFoundries and look into your own fab.

9/9/2010 2:21 PM EDT

Do you know how much it costs to have a fab? Try 1 Billion. Plus the headache of filling it to capacity. Real men go fabless because they are good in math...

If you are taping out a digital chip, there is no need to own a fab. TSMC, UMC or any other fab will do. Unless they have analog chips that need secret recipes to make it work...

Apple doesn't even have its own manufacturing. It all done in China by Foxconn. That's why it's so profitable.

Adam Smith in his 'Wealth of Nations' described precisely why this 'division of labor' can create wealth.

Asking Apple to have a fab will definitely cause both their profit, gross margin & stock to tank..

9/30/2010 3:52 PM EDT

Apple would benefit more from staying away from the Foundry business and negotiate with the current ones. There are some foundries to talk too and Samsung is not the only one ....

But Apple has already looked at second sources or else they would not be Apple ...... The negotiations are probably already at the end phase and pretty soon you would see the announcement.

9/30/2010 11:53 PM EDT

Yes ofcourse. They can easily acquire some fab firm. But do they want to enter that business?

10/1/2010 2:45 PM EDT

Apple seems like a popular subject on here. All it takes to get people talking is mention Apple in the headline!

10/20/2010 2:02 PM EDT

Apple should design their own IC's but, not necessarily own the fab house. I am most familiar with power converters bought from the far east which where huge, bulky, noisy modules no better than hobby junk from the 80's. How many millions of dollars due to lost sales on the mac7 because of these cheap things blasting noise? Apple could have it's own dedicated IC's designed into compact circuits with the technology of 2010.

2/16/2011 5:16 PM EST

"Logic fabs can cost $4 billion to maybe $10 billion to build out, depending on the tool sets, but Apple could swing the funding."

Perhaps so, given how much silicon volume Apple generates. But consider this:

* A fab would only handle one type of fabrication, for example the ARM SoC chips in CMOS. Apple has significant costs in FLASH and the displays as well, so all that capital would go to solving only one part of the total BoM.

* And all of that capital would be sunk into a single node, likely 28 or 22nm for a fab targeted to open in a year or two. It would take an even larger investment to convert to the next node. By using foundry partners, Apple can purchase latest node devices at market rates, without the capital outlay.

Apple has made some adept moves in this area, like switching from to Intel processors for Macintosh. This allows them to ride Intel's cost and performance curves without capital investment. It is unlikely they will suddenly decide their best use of capital is to build a fab!