Intel Delays Opening Fab Plant in Arizona
The new chip factory will be used for future technologies instead.
Is it a sign of the times? Intel Corp has reportedly put on hold the opening of a new chip manufacturing plant, "Fab 42," at its site in Chandler, Arizona. The facility, which was slated to allow Intel to create transistors using a 14 nanometer process, will remain closed for the foreseeable future. Other factories at the site will be upgraded instead.
The company said in 2011 that the Fab 42 project would cost an investment of over $5 billion USD. This is the very plant that President Barack Obama once held up as an example of U.S. manufacturing potential during his 2012 re-election campaign. Now the plant remains empty.
"The new construction is going to be left vacant for now and it will be targeted at future technologies," Intel spokesman Chuck Mulloy told Reuters. He informed PC World in an email that "we've been headed down this path for some time," and that construction of Fab 42 was completed on schedule.
Mulloy also told PC World that Intel will still deploy 14-nanometer technology in Arizona, but will do so in the existing Fab space at the site rather than use the new facility. As previously stated, the new space will be set aside for new technology, whatever that will be, as there's nothing in the new space save for heating and air conditioning; there are no manufacturing tools to be found.
"By running both 14 and 22 nm processes at the same time we get better utilization of our equipment," he said.
The news arrives after the IDC said that PC shipments took a 5.6 percent dive in the fourth quarter of 2013, with only 82.2 million units sold. For the full year, unit shipments declined 10 percent from 2012, a record drop due to changes in mobility and personal computing. Commercial purchases helped prevent a larger decline while the consumer side remained weak.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
-
vmem with AMD's new roadmap essentially announcing the end of CPU wars (it's been unofficially over for a while now), Intel's sole interest now lies in the mobile market and cost management. upgrading existing fabs instead of creating a new one is a good example of cost management... *sighs*Reply -
jimmysmitty With weak competition in the desktop and server market, Intel doesn't have to push as hard. They will use it when they need it, most likely to get a more advanced process up and running when they start their push into the UMD market more but for now they don't need to be able to FAB 14nm as no one is even close to beating them on process size and probably wont for some time.It is a bit sad as that's a lot of jobs that will go unfilled for some time instead of putting more money into the economy.Then again I would not be surprised if it was due to the local AZ government screwing it up somehow like Tucson screwed up and lost a huge facility from Raytheon. The city and state government for AZ tends to not work well with big businesses that will bring in a lot of jobs.Reply -
bharatwd Lack of CPU competition is stalling the market........which means slow rate of increased performance at higher costs.....i see the price of haswell is at 330$, the same price it was at launch.......sick stuff.............Reply -
lastxfirstx Jimmysmitty, you are correct that Tucson has made a number of bad moves where big business is concerned. However, Tucson is not Phoenix (Chandler in a suburb) and Phoenix has always been very business friendly. Probably the reason Intel located there. The problem with Tucson is only Democrats are elected and all races are determined in the primaries. That means you need to be far left and business unfriendly to be elected.Reply -
daekar Beyond the obvious end of the CPU wars, I wonder what the reasoning behind this is? I get the feeling this means something significant about Intel's future strategy, but I can't figure out what it is.Reply -
ingtar33 this just means intel doesn't think it needs to make as many desktop chips as it once did. that new factory was for desktop chips... and they're clearly aligning themselves to compete with ARM at the moment, now that AMD is no longer really all that competitive.Reply -
michel leclerc intel time table was screwed by relative ease of availibity of graphene ! if intel had ignored graphene as a material ,arm and its allies would have likely gained an advantage intel would have had a hard time to catch up! so intel played it safe and delayed (probably for a year to assess graphene potentiality as a material . I doubt they ll need more then a year tho !Reply -
obama_suks You guys are all clueless. This FAB was constructed ahead of time to beat EPA regulations that are coming down the pipeline that will destroy high-tech manufacturing in this country. How ironic is it that Obama came down to take credit for something his administration was only responsible for in the sense that they are making it so onerous to do business in the States that $5 billion dollar buildings are constructed in the desert to just sit there so that the current EPA regulations are grand-fathered in.It's the same thing with Obama taking credit for all of the new oil and gas deposits being mined on private land when they have done everything possible to prevent new deposits from being extracted on public/federal land.In both cases, Obama has done nothing but impede business and he tries to take all the credit for anyone succeeding or doing something, in-spite of, not because of, his administration.Reply -
somebodyspecial 12450148 said:With weak competition in the desktop and server market, Intel doesn't have to push as hard. They will use it when they need it, most likely to get a more advanced process up and running when they start their push into the UMD market more but for now they don't need to be able to FAB 14nm as no one is even close to beating them on process size and probably wont for some time.It is a bit sad as that's a lot of jobs that will go unfilled for some time instead of putting more money into the economy.Then again I would not be surprised if it was due to the local AZ government screwing it up somehow like Tucson screwed up and lost a huge facility from Raytheon. The city and state government for AZ tends to not work well with big businesses that will bring in a lot of jobs.
No, I think this is due to fabs not running at full capacity already (many posts around the web saying LOW capacity). They were building like they would own a portion of the mobile market, but instead chromebooks have taken 21% of their notebooks, and they have NONE of mobile. So fabs you already had are running low and no point in expanding to something that would just sit idle until you get 10%+ of the mobile market. Also chromebooks steal many parts from Intel (no cpu, chipset, boards many times, nics, etc). Intel has to push HARDER, they just can't without costing tons and having empty fabs.
http://intelstudios.edgesuite.net/im/2013/pdf/2013_IM_Smith.pdf
Intel says already 80% on current fabs from recent analyst meeting, and nothing new to fab until they get something in phones. You can't fab more cpu's to a shrinking market, you'll just end up writing them down later. They could delay tech and sell current gen longer but that won't start ARM's 64bit assault on your market despite AMD being weak. AMD doesn't count in the race now, it's Intel vs. the ARM-ada. ;) PC TAM down 10%, netbooks down 7% etc means IDLE fabs and those suck you dry. They are predicting 4x tablet growth for themselves so maybe AZ fires back up later this year. Their modem revenue dropped as the world moves to 4g and they're still working on that one (they are now in NV's shoes of 2013 etc). Sofia isn't coming until 2015 so won't be joining the phone race to fill more fab space this year. They should have pulled in a better modem like NV did. They will have a modem this year, Intel still won't (ramping h2 2014). Datacenter growth is really helping them mitigate PC slowdown but that doesn't help the fight with ARM vs x86.
The lack of production is why they are now going to product a few arm chips (of course ones that don't compete with them directly, not SOCS for phones/tablets etc I mean...they are other things).
Phones/tablets really took off 2011-2013 and they haven't hit 90% utilization since. So why do anything with the new fab? You can't fill the old ones as well as you used to, so just convert those to newer tech instead right (450 wafers will just make this all worse, as even more chips pop out of the same places if upgraded)? Makes sense. GM down 5% since the ARM war got ramped also. You can see AMD's 1999-2003 effect also when margins tanked last time by 14% :) Thank god for ARM and it's friends becoming Intel's 'NEW AMD" or we'd probably see a $200 range for junkers shortly and a 400+ range for high end again. Of course if Intel puts out something that takes 25% of the tablets and maybe 10% of phones we may see that anyway if they think they're gaining big time and will eventually make ARM guys like AMD today. I doubt this though :)
They seem to think they can increase the dividend, but I submit they can't afford the current one for more than a few more years (4-5), and when that day comes they have to lower it? BAM, down goes the stock price as investors get the fear of god in them...LOL. Running operations gets a lot tougher with all the things this hurts. Then everyone loses faith and more exit causing something like an AMD market value drop from ~8B to 3B.
Wintel is in trouble (which also means AMD since not on ARM yet and not for a while), I'm not the only one saying it:
http://www.thestreet.com/story/12234355/1/pieces-are-in-place-for-microsofts-death.html?cm_ven_int=obinsite
I'm not seeing death yet, but I can see both going to half their value easily in 5-10yrs (I'd maybe say 3-5 actually, but hedge my bet with 5-10...LOL). Hope that makes sense. Message seems clear me, the big picture says anyone not in ARM/Android is going to have problems. Intel obviously bleeds less than MS over this war.
Just got a great idea, maybe MS buys NV as defense? Might be a great move...LOL. They have far more cash than Intel to burn, and could buy them for $14-16B easily cash (MS has north of 70B last I checked - no brainer purchase for NV?). Suddenly they have a great SOC for chromebooks, nokia phones, RT crap, tablets etc and a company already making 500mil/yr not to mention all the IP they have. It gives them the next Xbox chips based on Denver rev3 or whatever too (xbox2 for 4K or something in 5yrs?). Hmmm...Great defensive move for either company but for intel I think it ends up costing too much and making it like an AMD->ATI purchase where they just can't afford it (could be wrong, MS seems better able to cash this out). Intel would likely have to borrow to get this done. Someone better make a move before NV gets too expensive and starts to dominate mobile gpus. I predict NV puts the hurt on Qcom+Intel over the next 5yrs. Which in turn means MS/Directx/games on DX also, which will also affect the consoles of course. GDC2013 in 2 months. MS purchase of NV allows them to get a little revenge on Intel for going android too...LOL. MS surely would have more options buying NV than just Nokia alone which was a waste IMHO. But the two combined and as a console leader, could slow the bleeding and maybe get some footholds of the future. Jen Hsun might not mind running the gaming division at MS (where he'd want to be CEO top dog at Intel which is why that probably never works, Intel hates him but needs him...LOL). Just food for thought, crap bouncing around my head today...ROFL.
Qcom earnings 29th. AMD shortly (which shouldn't be too bad as consoles are in there and sold well). Intel shortly too. Fun times to watch the show.
***runs to get popcorn*** ;) -
somebodyspecial http://www.itworld.com/data-center/352199/idle-intel-fabs-foretell-pc-declineReply
One more article showing Intel fabs at low utilization. They need to fab something more than just alcatel etc to get this back to 90+. Or the AZ fab is just pointless. I think they heavily overestimated their power to move to mobile and take on ARM.
"And even then, there's a chance Intel won't bounce back to its previous levels. Intel makes PC chips and the PC simply isn't where technology is at these days.
"This is the first time we're seeing the entire industry coming up fairly well, but the PC sector is not leading it. The PC is not driving the direction of technology anymore, it's not driving processor technology and it's certainly not the software magnet for technology anymore. It's a tablet and smartphone world," said McGregor."
So, MOBILE is leading the charge. 8 months later we're seeing this crap coming true and going forward looks even worse. While Intel can't up utilization, TSMC etc have no issues fabbing all the crap hurting Intel sales. How can you fix that without fabbing the enemy chips? Buying NV is one option, but a tough pill for all involved to swallow. Jen would want to run the show, nobody at Intel would like that...LOL. Too much hate.