Intel is routinely called a bellwether of the personal computer industry, and for good reason. Not only do they supply most of the world’s CPUs but they have consistently run a highly efficient organization that has achieved the kind of gross margins most other chip manufacturers can only dream of. So when Intel is doing well it’s often a sign of success for the wider PC industry, and when Intel begins to look uneasy it’s usually a sign of things to come.

This is perhaps no more true than right now, where the PC industry has been an extended slump. The hard drive shortage that initially pulled down the PC industry has since abated, but this has been replaced replaced by concerns over a slower economy, not to mention consumers shifting attention and money away from PCs and into tablets and smartphones.  Intel of course has exposure in both the PC and mobile markets, but being a relative newcomer to the latter they are still primarily vested in the PC market, and hence are exposed to any major market weaknesses.

This brings us to Intel Q3 earnings announcement, where we’re seeing those weaknesses take root in Intel’s earnings. For Q3 of 2012 Intel reported $13.5 billion in revenue with $3 billion in net income. This is by most standards a tidy sum – even when those standards are Intel standards – so it was by no means a bad quarter.

But when you start comparing this quarter to past quarters, an otherwise strong quarter stops looking quite as good. Compared to their record-setting performance last year in Q3 of 2011 Intel’s revenue is down 5% and their net income a much sharper 14.3%, offering a rather straightforward look at the weakness in the PC market. Meanwhile gross margins are Intel’s one bright spot for the moment, thanks to the fact that Intel’s gross margin having only fallen by a trivial amount in the last year; their margins have declined from 63.4% in Q3 of last year to 63.3% this year. Again this isn’t a bad quarter for Intel – it’s hard to understate just how good those gross margins are – but this is the significance of being a bellwether.  If even Intel is seeing a drop in net income, what does that mean for the industry as a whole?

Intel Q3 2012 Financial Results
  Q3'2012 Q2'2012 Q3'2011
Revenue $13.5B $13.5B $14.2B
Net Income $3B $2.8B $3.5B
Gross Margin 63.3% 63.4% 63.4%

The answer it seems may not be good. Intel’s Q3 results are only half of the news today; the other half of that news is Intel’s projections and planned actions for Q4. Intel has taken an unusually conservative stance on Q4, projecting that both revenue and their all-important gross margin will be down in Q4 of 2012 as compared to 2011. Specifically, Intel is projecting that their gross margin will be 57%, as compared to their incredible 65.5% gross margin from last year. On a quarterly or even yearly basis this would be a major shift for Intel, and it is a strong indicator that the weakness in the PC market will hurt Intel by more than just reducing their net income a bit like what happened in Q3.

This brings us to Intel’s unexpected revelation during their earnings call and the primary reason we're covering this: idle fabs and inventory buildups. Due to a buildup in inventory and a desire to prevent further buildup in what Intel is projecting to be a weak quarter, Intel will be taking the unusual step of letting quite a bit of fab capacity go idle. For Q4 Intel’s fab utilization will be sub-50%, with that being a combination of capacity idled to keep supply down and another fraction idled for the upgrade to 14nm. Because Intel would rather have too much capacity than too little they routinely take fab capacity idle for both production purposes and upgrade purposes – i.e. the 14nm downtime would need to happen regardless – but it’s very rare for Intel to let utilization fall below 50% like this.

Rolling this back into Intel’s finances, Intel’s incredible gross margins hinge on running their fabs at as close to full capacity as possible. So while idle fab capacity is still better than an inventory glut (or price cuts), it’s not without consequences. The consequences being Intel’s relatively low gross margin forecast for Q4, which if it comes true would be Intel’s lowest gross margin since 2009 at the end of the recession.

On that note, perhaps the cruelest part of this is that because Intel’s overcapacity appears to be at 22nm – Intel’s standing inventory is nearly 70% Ivy Bridge – Intel’s options are to produce Ivy Bridge or go idle. Atom is still built on Intel’s older 32nm process, so even if Intel is in a position to capitalize on the strong mobile market they can’t shift from producing PC CPUs to mobile SoCs. In a situation like this Intel’s accelerated Atom roadmap can’t come soon enough.

Moving on, the real wild card for the moment is Windows 8 and what impact it will have on PC sales. The Windows-induced Osborne effect is practically a tradition in the PC space so it’s expected to see PC sales weaken ahead of a new version of Windows. The difference this time is that the weakness in the PC space isn’t just delayed demand based on Windows, but also due to the aforementioned economic issues and competition from mobile devices. Intel certainly isn’t expecting Windows 8 to reverse PC sales on its own, but just how well it is received will definitely have an impact on Intel’s Q4.

Ultimately the Intel story will be the template for the rest of the PC industry. Regardless of demand for Windows 8 it’s almost certainly going to be a soft quarter for the industry, and unfortunately few companies are in a position to weather it as well as Intel. For those like Asus who have a significant presence to the mobile market they can look forward to a strong mobile market offsetting PC declines, while other manufactures like HP and Dell will face the full brunt of any changes in the PC market, be it good or bad. Even with the hard drive shortage the PC market has enjoyed a good run for the last few years, but if Intel’s Q3 results and Q4 predictions are anything to go by a cyclical slump may quickly be approaching.

Comments Locked


View All Comments

  • Shadowmaster625 - Wednesday, October 17, 2012 - link

    Rather than use that fab capacity to make a tablet chip that can compete on price, they just idle their fabs so that they can continue to force ivy bridge slate prices to remain at the $800-$1000 mark so they can continue to be eviscerated by crapple. Smart, smart strategy! If I can get an atom tablet for $450 there aint no reason in hell I shouldnt be able to get a celeron ULV tablet for $50 more. Intel has been selling 10W celeron/pentiums for 5 years now, for $78 or even lower.
  • tipoo - Wednesday, October 17, 2012 - link

    It takes time to move any processor to a new fabrication process, the article says that most of the idle fabs are 22nm, which neither the regular or smartphone Atom is at yet.
  • 0ldman79 - Wednesday, October 17, 2012 - link


    Odds are Intel has been working on the Atom's move to the 22nm process for a while now, they're just not there yet.
  • DanNeely - Wednesday, October 17, 2012 - link

    It's not just a case of moving the atom to a new process. The atom is built on a different process than Intel's other CPUs. The Atom is built on a low power process, everything else from the Celeron to the Xeon uses a high performance process. The performance process is about 10% faster at the cost of higher leakage current, etc (don't recall how much this hurts).

    Intel's low power process has always lagged behind their high power one by about a year; I believe the plan is to catch up so that beginning with 14nm the low and high power processes initial production dates are in sync with each other.
  • Homeles - Wednesday, October 17, 2012 - link

    You don't understand how processor design works and the amount of time involved. Please educate yourself before making inane comments.
  • semiconshawn - Wednesday, October 17, 2012 - link

    The just need to foundry for Apple. Money and fab cap solved. Intel won't foundry but it would be cool to see 22nm and below arm chips a year or two early.
  • menting - Wednesday, October 17, 2012 - link

    Personally, I think that's out of the question.
    Apple is already too big for Intel's liking. Maybe for some other companies in the future, but I don't see it for Apple.
    Unless Apple is using some mobile chip from Intel,I don't see any way Intel will foundry for Apple.
  • tipoo - Wednesday, October 17, 2012 - link

    On the other hand Intel did talk about starting to manufacture other peoples processors earlier this year I think, and a customer is a customer. Look at Apple and Samsung, still leaning heavily on each other despite everything. A company can't say no to money.
  • amdwilliam1985 - Wednesday, October 17, 2012 - link

    I read it some where that Apple and Samsung is breaking up in chip manufacturing. So Apple needs to find itself a new foundary to work with, maybe intel is the key, but I highly doubt it. Just like some other commentor mentioned before, Apple is already too big for intel's like. Plus some other rumor about Apple moving away from intel's x86 chips, that will definitely stir up more bad feelings between the 2.

    I think one of the key factor here is Steve J, Tim Cook is no Steve J, he can't fight with a company and then makeup with them instantly. Since Steve J(please clone him and bring him back, lol) is no more, I'm interest in what Apple's move next.

    The only reasonable alternative to Samsung foundary is intel.

    I'm also interest in intel getting into the ultra low power cpu, I heard the intel-powered Razr M is not bad. I'm interesting in more of their offerings in 2013.
  • DanNeely - Wednesday, October 17, 2012 - link

    I'm not sure Intel is the only alternative (although with their process lead they're probably the best one if willing). Rumors about a deal with TSMC have been floating around recently. Global Foundries could certainly use the business; and Apple's both shown a willingness to dump its money into component makers to accelerate their production of new devices and one of the few companies that actually has deep enough pockets to conceivably pull them out of their tailspin. Beyond that IBM still has a CPU foundry; although I don't know if they have a low power process or not.

Log in

Don't have an account? Sign up now