Qualcomm's current supply problems come as part of a fabless strategy. You can't have it both ways. If you don't own a wafer fab, then you can not control your supply of chips. If you do own a fab then you have to pay for it when you have excess and, or, obsolete fab capacity. It's going to be expensive either way. In the last several years being fabless has been very good to Qualcomm. It may now change for the next several years. Right now it seems IDMs like Intel and Samsung have the advantage.
Qualcomm should ask SanDisk or Amkor, two examples that fabless and OSAT tried of getting into the business that they are not familiar with.
This may be just a bluffing to tell foundries to get their act together.
I think the best bet for Qualcomm is to invest in someone with the capability of 28nm and beyond already, to get their quaranteed capacity. But there are a lot more business consideration and technical, for example, Samsung may be good candidate since QCT still a major supplier to Samsung's cell phone, but not to put all the eggs in basket as one day Samsung may become a competitor as in Apple's case.
There's nothing wrong with UMC or GlobalFoundries and -- as i discuss in an analysis -- I think GlobalFoundries could be a major supplier to Qualcomm.
However, the days when you could easily move an IC design from one foundry to another are long gone.
So are the technologies incompatible...well both foundries are offering CMOS.
But they are different CMOS processes, with different tweaks, design rules, checks and sequences. There is considerable foundry supplied IP required to make a design work and different foundries have different IP sets and design rules associated with those IPs.
The result has been that fabless chip companies have tended, even where they work with multiple foundries, to only bring a given design up at one foundry. To run genuine multivendor second sourcing is generally thought to be too expensive.
And at the leading-edge you would have to wait for a second foundry to match the first's capability. Waiting is not a good idea in consumer markets.
However, when you are single-sourced on a leading-edge process and you do not get the volume of chips you desire you start to think "How could i get favorable treatment? I want my allocation of chips and my rival's allocation."
Peter, I would agree that moving from one foundry to another is no trivial task! Qualcomm would be wise to invest in a part of a fab as a second source backup and to allow for upside production. I wonder if there is an available existing fab that would be reasonably compatible (or could be tweaked to be) with their existing foundry process?
A far cheaper way for fabless companies is to develop some independent DFM expertise and make their layout manufacturable everywhere.
The rules are more or less the same because the fabrication equipment are the same and the EDA software are the same. Fabless companies can build or uses or transform an existing standard organization to do it (Si2, Sematech, etc.).
For now, we are seeing the effects of de facto foreign monopoly. A monopoly helped by the fabless companies. Consumers will pay more.
Agree but also disagree... Fab processes are not as simple as any typical manufacturing or assembly line. Layout is important but there is tons more into it other than layout.
I do not see it as de facto foreign monopoly. Not only the R&D but also the FAB processes are the key assets of IDM's like Intel, Samsung, Miron, IBM, TI,...and the like.
For the big-Q, even if they start investing on their own capacity now, FAB mfg is not like making Domino's pizza that all you need is the recipe and ovens. You cannot just buy tools, lay them out in the FAB and start manufacturing wafers
Used to work like that until 40nm when only variables are stress engineering and gate oxide but not anymore.. The differences of advanced technologies are so subtle and is reached to the point that even Qualcomm do not have enough resources to make their design multi-sourceable and EDA tools are not sufficient(not upto par with technology changes) to address these issues.
Still, it is cheaper than building a fab.
For a company that couldn't tell oxidation from rust, the cost for IP would make a lot of lawyer very happy.
Anyway, it is a good joke. I can hear the roaring laugher coming from Shinchu.
Considering the rising cost and unclear roadmap to future CMOS, it doesn't make much sense to build their own fab at this point. Especially they are competing Intel and Samsung in the same market mobile processors. If they were using 90nm technology, then maybe. The statement from Qualcomm is directed at TSMC and pressure them to deliver on time.
@Howard, I totally agree with you. It doesnt make sense to build own fab especially when the technology is evolving so rapidly. Let's not forget TI which once held both analog and digital fab's closed down its digital fab's and just retained all the analog fabs.
The semiconductor industry is very complex. A lot of IP is involved. The fact that the latest S4 processors are 28 nm based rule out the multi-vendor option. I think if Qualcomm is talking seriously the next logical step is to go for a fab. the question is when is the right time?
Also, this would only be to address the shortage of Snapdragon processor? How about other chips from the fabless giant?
I have said this before. No one can discount the impact of fab in Intel's success. Intel continues to win because its innovations are double-edged. They have process innovation besides all they do with circuits and architecture. If Qualcomm go do same, they may get it right. Yet, it is very expensive and risky.
If Intel fully succeeds in their own terms they will (yet again) drop all foundry business.
Their statement that fabless is going nowhere can (should?) be read as an indication that they won't take support of fabless companies seriously.
@goafrit, No doubt having its own fab is helping Intel to succeed but Intel has built the expertise to build and maintain the fabs over years. It would be a herculian task for the qualcomm to build and maintain its own fabs at this stage when technolgy is shrinking so fast.
I agree, it'd be crazy for Qualcomm to jump into a bleeding edge logic fab - but hey, it was crazy for ATIC to create Global Foundries and jump into the foundry business with no pre-existing expertise. But they did it anyhow, and it seems like they're able to tread water. So maybe Qualcomm will pop a crazy pill one morning and take a go at it.
The only "jump" at Qualcomm is the suicide that happened last month on June 11. Well another one 2 years ago, but who's really counting now right? It's no mystery that the June 11 incident didn't get reported in ANY news outlets, in any sort of media. When there's a suicide in the city, it gets reported. Question is how much money did Qualcomm pay off to keep this a hush hush, just so that their stocks don't get a dive? Very unethical of a company of that size. The fact that everything's so secretive makes employees working there want to spread the word out even more.
I had assumed that the reason that Qualcomm was not getting volume from TSMC was that they had tried to minimise financial risk by not pre-booking (sufficient?) capacity. In the event that capacity was under-utilised, this would in the past have allowed Qualcomm to negotiate better prices (it's not really the case if you are sole-sourced, however).
If my assumption is correct, this represents at least two major failings in Qualcomm's purchasing strategy - both a misreading of the semiconductor demand/supply cycle, and failure to recognise the implications of single sourcing.
I can see no reason that self-supplying would compensate for incorrect prediction (more the reverse, in fact).
If there were any business sense in the idea of floating an own-fab it would be as a negotiating ploy - however, the potential suppliers have too much information for it to work.
I believe the only possibly "benefit" of Qualcomm's public stance is to avoid excessive fluctuation of the share price - and even this may well misfire.
The issue of pre-booked capacity is also inextricably linked to yield. Which then begs the question as to whether low yields are the due to poor design (the fabless) or the process (the foundry).
Qualcomm may be arguing that at 80 percent yield they would have had enough wafers but at 20 percent yield they only have one quarter of what they need.
TSMC's response may well be that at this stage in a roll out lower yields are quite normal and that Qualcomm should either have bought more wafers...or should make its customers wait for their chips.
As such, mentioning alternative manufacturing strategies may be a way of putting pressure on to TSMC...a coded way of saying "You are my supplier and are meant to be trying to keep me happy. I am not happy and have to think about my future here."
How would yield be better if you own the fab? Your own learning curve would be at best at par or worse than TSMC at each nodes. If you do the math right, it does not even make sense for a company like Qualcomm to own a fab. They always had a option to purchase additional capacity ahead of time with take or pay option in their supply agreement if they planned it right. The problem is that they never execised its option until it's too late. To me it's a Qualcomm's indecision and management issue and not able to plan their demand picture correctly. They're simply shifting investors' focus from inability to manage their demand to supply issue. There is no supply issue here if they planned it ahead. Now they are going around every foundries and trying to drum up the capacity - which in turn is a wrong move.
I think that if QCT had their own fab, they'd enjoy the same advantage that Intel does - that is, optimizing your process for one design, rather than just using an SRAM or something and then having to tweak your devices for each customer that comes in after development is done. But your larger point is right - management motives are usually less than transparent.
Well, it may figure that a fabless company, even Qualcomm, may be averse to facing the fab-intensive risk and complexities of scaling to the next node, to the point of not wanting to face it earlier than necessary. So maybe it is not the 'fabless model' that is the trouble, but the fabless mindset.
Those who were paying attention when DFM had its 15 minutes of fame at DAC and VCs should have seen that the bandaid solutions which worked at 65 and 45nm wouldn't work at sub-20nm IF EUV didn't make it. And those who remember that EUV is still x-ray should have seen that EUV wouldn't make it. The only way to build 10nm logic devices will be with completely seamless integration of design and manufacturing. For Intel and other such IDMs, this is second nature. Fabless/foundry will require a new model. Perhaps reintegration? Buying a fab? Renting a fab line? Otherwise, IDMs rule.
Even the companies that have fabs use foundries. It's an easier and quicker way to increase capacity and allows them to decrease capacity without running their own fabs at less than full capacity. It's a nice hypothesis, but the actual results aren't that easy. For a fabless company to build their own fab is a huge task. The time and cost up front is huge. I don't see that happening for Qualcomm.
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...