Intel Contemplates Outsourcing Advanced Production, Upending Oregon's Central Role (oregonlive.com) 108
According to The Oregonian, Intel is "openly flirting with the notion of moving leading-edge production from Oregon to Asia and hiring one of its top rivals to make Intel's most advanced chips." The decision is likely in January. From the report: It's a momentous choice that follows a string of manufacturing setbacks at the Ronler Acres campus near Hillsboro Stadium, failures that have cost Intel its cherished leadership in semiconductor technology -- perhaps forever. Outsourcing wouldn't shutter Intel's Oregon factories or close down its Hillsboro research labs. The company says it's committed to maintaining its advanced research and retaining internal production capacity. It's continuing a massive expansion of its D1X factory in Hillsboro.
In time, though, Oregon's central role in Intel's technology would almost surely erode if the company cedes manufacturing leadership to rivals overseas. Chip industry analyst Dan Hutcheson of VLSI Research believes that transition could render Oregon "irrelevant" if Intel gradually shifts away from integrated research and manufacturing. "Companies say they're making a transition. What they find is they're stepping off a cliff," Hutcheson said. "They're going down a road that you can't easily go back on."
CEO Bob Swan told Wall Street analysts on a conference call earlier this month that it may outsource advanced production to its rivals -- he named Taiwan Semiconductor Manufacturing Co., specifically -- to ensure "a predictable cadence of leadership products." Swan told investors to expect a decision by late January. Intel already outsources as much as a fifth of its production but has kept the leading edge in Oregon. And whatever it decides on outsourcing, Swan said Intel will maintain its advanced research -- science performed in Hillsboro -- which he described as "a powerful force in creating future differentiation for our products." On this month's analyst call, Swan said Intel believes it can have it both ways -- sending advanced production overseas while retaining internal production for components and older products that don't require the most sophisticated technology. And Swan said Intel believes it could restore advanced manufacturing to its own factories sometime in the future, if it chooses to.
In time, though, Oregon's central role in Intel's technology would almost surely erode if the company cedes manufacturing leadership to rivals overseas. Chip industry analyst Dan Hutcheson of VLSI Research believes that transition could render Oregon "irrelevant" if Intel gradually shifts away from integrated research and manufacturing. "Companies say they're making a transition. What they find is they're stepping off a cliff," Hutcheson said. "They're going down a road that you can't easily go back on."
CEO Bob Swan told Wall Street analysts on a conference call earlier this month that it may outsource advanced production to its rivals -- he named Taiwan Semiconductor Manufacturing Co., specifically -- to ensure "a predictable cadence of leadership products." Swan told investors to expect a decision by late January. Intel already outsources as much as a fifth of its production but has kept the leading edge in Oregon. And whatever it decides on outsourcing, Swan said Intel will maintain its advanced research -- science performed in Hillsboro -- which he described as "a powerful force in creating future differentiation for our products." On this month's analyst call, Swan said Intel believes it can have it both ways -- sending advanced production overseas while retaining internal production for components and older products that don't require the most sophisticated technology. And Swan said Intel believes it could restore advanced manufacturing to its own factories sometime in the future, if it chooses to.
Re: (Score:2)
Just wait until Tuesday next week, they'll be all Apple-only and Xbox-only news. And then on Thursday they'll be nearly all PS5-only news.
Man, next week is going to be busy.
Re: (Score:2)
When do we learn Mandarin?
Um, seven years ago. Where have you been?
Re: (Score:2)
Um, 14 years ago, where have you been?
Re: (Score:3)
So if Biden wins China designs our military's CPU's?
1. TSMC is not Chinese. Hint: Guess what the T stands for.
2. TSMC does not design chips. They fab them.
Re:Decision in January (Score:4, Insightful)
Re: (Score:3)
I wouldn't rule out an invasion attempt within next 5 years.
Go watch "Saving Private Ryan".
That is what an invasion attempt looks like when you are facing 10% of the enemy (90% of the German army was in Russia) and you have 100 times as much amphibious shipping as China has.
There is no way that China is going to have the ability to launch a cross-strait invasion of Taiwan within 5 years.
They Wouldn't Have To (Score:3)
All they would need to do is throw a naval ring around Taiwan and intercept and/or block everything attempting to enter or leave Taiwan by sea... and the same by air. The country is already close to completing (or have completed) Shuimen Air Base [wikipedia.org], and have massively extended another near Xiapu [defensenews.com], which is just 160 miles from Taiwan.
If China decided to annexe Taiwan in the same way that Russia annexed Crimea, the
Re: (Score:2)
I strongly suspect even a Biden administration would initiate an American naval response if China attempted to annex Taiwan. Taiwan is to integral to our own economy to allow such a thing, and it would be politically impossible to continue pretending China is not an aggressive expansionist and existential threat to America.
China's policy toward both Taiwan and us is one of boiling the frog. They keep turning the heat up little by little so the denialists can pretend its not happen and the rest of us get to
Re: (Score:2)
I wouldn't rule out an invasion attempt within next 5 years.
Go watch "Saving Private Ryan".
That is what an invasion attempt looks like when you are facing 10% of the enemy (90% of the German army was in Russia) and you have 100 times as much amphibious shipping as China has.
There is no way that China is going to have the ability to launch a cross-strait invasion of Taiwan within 5 years.
Sure they could. It would be risky but 15 neutron weapons over taiwanese military bases results in unconditional surrender in 72 hours.
Re: (Score:3)
TSMC needs to diversify geopolitically. If they start opening fabs in other countries - like the United States - then their exposure to Chinese influence is reduced.
Re: (Score:2)
Alternatively they can acquire and refit existing fabs. The US has a number of them.
Re: (Score:2)
It is generally not worth refitting an existing fab to a new technology node. Fab purchases are usually based on developing a new type of product at the same node.
Re: (Score:2)
Since when? Intel and TSMC refit their own fabs on a regular basis. It's not like you throw out the facility when you're done with the node. Do you have any idea how often the GF fab at Fishkill as been refitted?
Re: (Score:2)
Intel will either look for new products to fill the old lines or convert the space to assembly and test.
GF is selling the old lines at East Fishkill because there is another company who make discrete products at the same nodes. They built Fab8 Malta to do 14nm and 7nm. Fab9 is still running the old processes.
Re: (Score:2)
Malta will likely never produce 7nm.
Fab 10/East Fishkill has been sold to ON Semiconductor; regardless, it was one of the fabs they acquired from IBM, and it was fitted for numerous processes. The most-recent process IBM manufactured independently at Fishkill was (if I recall correctly) whatever process they used for POWER8. 22HP? In any case it has since been refitted for 14nm. Or they added 14nm lines anyway.
Intel, meanwhile, has actually done refits from 14nm to 10nm and then back to 14nm in some of t
Re: (Score:2)
Texas. Duh.
Re:Decision in January (Score:5, Funny)
No, seriously, do you think there is even a small chance that Biden would do something about this and Trump would not?
Biden certainly isn't up to the task of fixing Intel's manufacturing process to be competitive with global vendors. He's so old that he still thinks that three-micron lithography is impressive.
Trump, OTOH, is an extreme-ultraviolet genius: Just look at his tan!
He's got a beautiful plan to get Intel back on top of the nanometer treadmill, winning bigly! His plan will be out in just two weeks!
Re: (Score:1)
Just ask Foxconnsin.
Once manufacturing leaves, it has gone.
Re: (Score:1)
True, True.
Re: (Score:2)
Apparently you haven't seen the advanced LCD manufacturing plant they built here in Wisconsin!
Trump even started digging the hole himself!
Re: (Score:3)
He's got a beautiful plan to get Intel back on top of the nanometer treadmill, winning bigly! His plan will be out in just two weeks!
I think you mean his plan is just around the corner. Rounding the final turn. Any day now, like a miracle. The end is in sight.
It's actually depressing to think that the China trade war has forced them to accelerate work on developing the very fabs that Intel may need in the future in order to remain competitive.
Re: (Score:2)
Re: (Score:2)
It isn't complicated. Get existing semicon fabricator to lease/sell fabs to TSMC in exchange for a cross-licensing deal. GlobalFoundries and (guffaw) Intel are both in a position to take advantage of such a deal. If TSMC can get significant fab footprint in North America, China can't bully them. As much.
Re: (Score:1)
ultraviolet genius
this one is good -- elegant & effective on its own.
but when it's in this sentence:
Trump, OTOH, is an extreme-ultraviolet genius: Just look at his tan!
the whole thing tastes like an over-processed, salty mess.
you want fries with that?
Re: (Score:2)
You'll need EUV wavelengths to process features at ever-shrinking scales. Probably also to generate that particular skin hue.
And, the accountants again take a company down.... (Score:3, Insightful)
It's not the accountant getting their asses kicked (Score:3)
It's not the accountants who are unable to figure out the physics of reliably fabricating wafers with the advanced processes TSMC can. TSMC produced 10nm wafers in 2016, Intel followed two years later in 2018.
It's not the accountants getting their butt kicked in benchmarks:
https://www.cpubenchmark.net/h... [cpubenchmark.net] (and it's even worse if you click on performance per dollar).
The accountants are the ones to crunch the numbers and figure out just how fucked they are.
Re: (Score:2)
A fabless Intel is a slightly less fucked Intel. The sale of the Intel name will garner more dollars than the sale of their manufacturing division (sort of like what happened to AT&T) but it is the fabrication step that makes most of the profit (not at all like what happened with AT&T.)
The company faces a decision. Keep the fabs and fall into obscurity but reliably make at least some money in the future, or sell the fabs and event
Re: (Score:3)
Right, Intel has been struggling with its processes for years now and has already passed the point where they are competitive. Stuff like AVX512 shows just how desperate they are.
Re: (Score:3, Interesting)
This is a long time coming and it didn't start in 2016 or 10nm. This started a good decade or more ago.
The company lost it's hear and soul. Intel is a manufacturing company, but lost it's way.
1) Poor non technical leadership. This is where the fall started. It's hard to think for most of the early to mid years of Intel, the company was run by actual technologists. People who developed stuff that had never been done before, ran the company. As soon as this was lost the fall started.
2) Focus on stupid
Because 369 million is more than 70? (Score:2)
Just curious, when is the last time you bought a new desktop PC? It's been probably seven or eight years for me.
> 2) Focus on stupid products. Why the hell did you ever try to make a phone, or chips for a phone, or anything to do with a phone?
Intel makes computer chips, processors. Last year, 369 million computers were sold in the portable / phone format, 70 million in the desktop / laptop format.
For a major processor company, ignoring the vast majority of the proccesor market would be dereliction of
Re: (Score:2)
Re: (Score:2)
Simple. Desktop Processors at the time had 50-80% profit margins. Small ARM processors have 10-20% profit. Very small margin. Server Processors were way off the charts.
Re: (Score:2)
Have we already forgotten that the catastrophic Intel 14nm launch was under Bryan Krzanich? And that the Pentium IV "NetBurst" bullshit was under Craig Barrett?
This could be just the thing but it won't be. (Score:5, Interesting)
It's certainly clear that Intel is under intense pressure to produce results and they are currently not up to the task. Time is really what they are buying by offloading manufacturing. If they play their cards right, they could use this as an opportunity to leapfrog existing fabrication tech by deeply investing in a diverse array of emerging and undeveloped fabrication technologies with the bet that at least one will pay off.
Frankly, leadership at Intel has always been about quarterly results and making incremental improvements not about the development of actual technologies, so there is little hope of long-term thinking from upper management.
Re: (Score:2)
It's certainly clear that Intel is under intense pressure to produce results and they are currently not up to the task. Time is really what they are buying by offloading manufacturing. If they play their cards right, they could use this as an opportunity to leapfrog existing fabrication tech by deeply investing in a diverse array of emerging and undeveloped fabrication technologies with the bet that at least one will pay off.
No offense, but that is one big load of goggly-gook..and frankly, a load of crap. The have screwed up consistently for the last few years, and now they think this will dig them out of the hole they put themselves in....and it will only come back to bite them in their butt. Just my $0.02.
Re: (Score:2)
> hat Intel is under intense pressure to produce results and they are currently not up to the task.
Indeed.
> leadership at Intel has *always* been about quarterly results and making incremental improvements not about the development of actual technologies [emphasis added]
Hmm Intel has never developed any technologies?
Re: (Score:2)
Hmm Intel has never developed any technologies?
Not at all. Intel has developed tons of tech but for management it was always just a means to an end. They have no passion for the work, it's just numbers which is reflected by their constant and unending anti-competitive behavior for the last 30+ years.
Re: (Score:2)
Probably wouldn't hurt to be able to watch someone who knows how to make it work either.
Maybe get a tour of a working euv fab
Re: (Score:3)
Intel should outsource Bob Swan. (Score:1)
MIGA! (Score:5, Interesting)
Intel has recently sold its memory division, to "focus on core competencies". Now it's basically selling off its core competency. I guess we're in for a couple of very profitable quarters and several years of decline after that. RIP.
Re: (Score:2)
Speaking from experience... (Score:5, Insightful)
I've heard the same tale before.
Nokia and RIM used to be the kings of the (smart)phone world, but they become complacent and others caught up. The old guard couldn't catch up, and so the fat-trimming starts: "We will outsource the highest end production and design. Don't worry we will still make the mid and low end devices in-house." 5 years down the road, everyone else also caught up and now no one wants to buy the in-house products anymore. So the entire manufacturing division gets laid off and what's remain of production is outsourced to the lowest bidder.
Re: (Score:2)
Nokia also outsourced its leadership position to Microsoft's trojan horse Steven Elop, who in turn run the company into the ground
Re: (Score:2)
Re: Speaking from experience... (Score:2)
Re: (Score:2)
Comment removed (Score:5, Interesting)
Maybe trying to finagle more Oregon tax breaks (Score:3)
This could just be a negotiating tactic. Intel could also move to Texas if they wanted space and lower taxes.
Re:Maybe trying to finagle more Oregon tax breaks (Score:5, Funny)
It always amazes me that a chip fab, which uses 2 to 4 million gallons of ultrapure water per day would move to a desert.
Why Intel, why?
Wouldn't one of the great lakes states make more sense?
Re: (Score:3, Interesting)
1 acre-foot/year is ~892.74 gallons/day; 9 million acre-feet/year is, unless I'm slipping a zero somewhere, a bit over 8,000,000,000 gallons/day.
Re: (Score:2)
These are the corporate tax rates in the states surrounding the Great Lakes, along with a comparison to CA and TX (from https://taxfoundation.org/stat... [taxfoundation.org]):
In general, tax rates for most states that charge corporate income tax are around 5-7%. So I'm not really seeing "treatment of highly profitable companies as bottomless founts [sic] of tax revenue".
Re: (Score:2)
What talent would trade the West coast for the miserable winters of of the Great Lakes region?
Re: (Score:1)
Talent that cares about things like clean air, better schools, better quality life by just about any measure. I'm out of Minnesota, if there's a top 50 states list for something or another, chances are Minnesota's in the top 5.
Talent that cares about things like affording a place to live in a decent neighborhood. Talent that cares about spending time in a place that isn't quite so crowded or politically radical. Mind you, I've enjoyed traveling to California dozens of times over the years for one thing or a
Re:Maybe trying to finagle more Oregon tax breaks (Score:5, Interesting)
This could just be a negotiating tactic. Intel could also move to Texas if they wanted space and lower taxes.
No, it's real. Their floating a trial balloon, but it's obvious where they're going and why.
The people that founded Intel and knew how (and wanted) to operate an advanced semiconductor manufacturing business are gone. Now you have management and finance guys like Bob Swan calling the shots. He's been bouncing around the executive suites of various corporate zombies his whole life. He has no interest in spending time with his engineers and the feeling is doubtless mutual. He's a big wheel that wants to be cozy with the hedge fund guys and big hitters in government. The sooner he can shed all that nerdy foundry junk the better.
It's really that basic. I know it seems simple minded to the point of absurdity, but that's what is actually happening. Swan and the horde of finance people that are employed to to affirm his decisions have no interest in and find technical stuff frustrating and opaque. So they convince themselves that isolating the business from the vagaries of engineering is the correct path forward. They dream of reducing the hard work of technical problem solving and the tedious task of herding geeks to a contract negotiation.
Intel has died. Going forward it's a brand and an IP trading company.
Re: (Score:3)
I kind of agree but Intel has been trying to catch up for a while now and has little to show for it. Their EUV team is just not good. They are now getting laughed at as a company. Frankly even an engineer in charge would start doing desperate things at this point.
Re: (Score:2)
Re: (Score:2)
That isn't even true. AMD is making leaps and bounds with x86. They actively ditched their in-house ARM design - K12 - which frustrated Jim Keller enough that he left AMD. He eventually wound up at Intel and allegedly left for family reasons, though it's interesting that he got out before the power struggle between Bob Swan and Murthy resulted in Murthy being fired.
In any case, don't ascribe Intel's failures to an ISA. Their problems are with their fabs, not their designs (per se).
Re: (Score:2)
Re: (Score:2)
Insecure compared to what? You aren't even seeing what they've got working in simulation. Intel's fabs are holding back their designs, period. Golden Cove should be out right now without many (or any) of the security problems you're referencing above. So where is Golden Cove? Held over until 2021, maybe 2022 (Alder Lake). Willow Cove only exists as a 4c laptop CPU. Sunny Cove . . . 4c laptop CPU. Do you know why? It's simple: Intel's 10nm process went boom. And their 7nm process is delayed.
Re: (Score:2)
Agree'd x86 / x86_64 isnt a problem and it isn't the problem. Its not like the ISA isnt mostly anyway.
There real problems are the existing x86 market expects high single thread performance but the work loads of today's systems are evolving. You can't really go to market with x86 that runs like an ARM but there is wide space of consumer and server products where once x86 was about the only realistic option but can be driven by a ARM SOC. The low performance market for x86 could be recovered but to do it you'
Re: (Score:2)
x86 still dominates in the server space. AMD stands ready to snap up all that market share. ARM in the server room is dying before it got started. Only Amazon's Graviton2 is having an impact. Marvell killed ThunderX3, Ampere's offerings are MIA, and Huawei's Kunpeng 920 is a). embargoed and b). underwhelming.
Server/cloud is where the money's at, and x86 is still raking it in there.
Intel is screwed........ (Score:5, Insightful)
Intel is screwed, they can't engineer or manufacture a high end CPU these days without a flurry of problems.
Wouldn't it be funny if Ford had trouble making cars and called GM up and asked if they could manufacture the F-150 and Ford Mustang?
When it comes to this, you are screwed as a company.
I feel sorry for the once great Intel. They missed the boat on mobile CPUs and even lost the premier position with Apple. Intel is on a roll, but it is all downhill.
Re: (Score:3)
It's hard to feel sorry for them since they are a corporation, and a sleazy one.
I feel sorry for the employees who show up and try.
I don't feel sorry for the shareholders who bankrolled and profited from their fraud.
Intel screwed their own pooch with mobile CPUs (they bought an ARM design and then failed to make it low power, even though they had an unmistakable lead in process technology at the time) and with Apple (SPECTRE, MELTDOWN...)
You can pin their problems directly on their senior management.
Re: (Score:2)
Re: Intel is screwed........ (Score:2)
Intel literally knew that they were playing fast and loose with security. They were warned. It was deliberate.
Re: (Score:2)
Their chip design is doing just fine. Ice Lake and Tiger Lake may not be complete ground-up redesigns like Zen, but Intel already had a more solid foundation to build on The new chips, based on the Sunny Cove microarchitecture, achieve significant IPC improvements over Skylake and take care of the Spectre and Meltdown vulnerabilities in hardware.
But the potential for those designs is being held back by the company's inability to build competitive fabs for them. They haven't been able to get the clock rates
What AMD did (Score:3)
When in a similar situation, AMD spun off its manufacturing and became a fabless company. That, at least, is consistent with the general trend of stratification in the industry, and from what I have seen from the inside of Intel, the fab is 352 organizational layers away from anyone doing design anyhow, so what value is there in such vertical integration anymore?
Intel management seems to want to have it both ways here: getting access to capacity at the latest nodes, without having to manage the risk of building it out. The result to me looks like a half-measure more likely to fail than either committing to vertical integration or going fabless.
Re: (Score:2)
Due to the facts of this kind of manufacturing, they end up with "outdated" fabs that have nothing better to do than make ram chips, flash chips, microcontrollers, custom asics, and so on.
Intels problem is that they cant keep those outdated fabs running 24/7/365 like a rent-a-fab can, and none of Intels competition in any of those markets (ram, flash, etc) would ever rent time and hand over their non-patented secrets at the same time. In essence, because
Re: (Score:3, Informative)
When in a similar situation, AMD spun off its manufacturing and became a fabless company.
They were never in similar situations, even when AMD still had its fabs.
In any case, AMD has demonstrated what can be achieved under pressure with limited resources; Intel's shown us what MBA's are capable of when under no pressure and equipped with (practically unlimited) resources.
Corporate America supporting America (Score:3, Insightful)
Intel moving advanced production offshore (Score:3, Interesting)
Re: (Score:2)
Lately nVidia has been using Samsung, but otherwise your point still stands. TSMC has publicly stated that they can't and won't take on all of Intel's chip volume. So something else major would have to happen for TSMC to reverse course.
Handing it to China (Score:3)
And what happens when China finally tightens the leash on Taiwan?
Re: (Score:2)
And what happens when China finally tightens the leash on Taiwan?
Exactly. Intel is just handing its advanced technology over to the Chinese. Taiwan is just chock full of mainland China spies and sympathizers. And it may even be invaded successfully if it doesn't eventually just hand itself over to China due to its intense program of corrupting and influencing the Taiwanese to do so. Taiwan should have a policy of blowing up all its chip research and production facilities in the event of an invasion by China. Also, Intel may be actually increasing the danger to Taiwan by
Re: (Score:1)
Re: (Score:1)
This is what happens when (Score:3, Insightful)
Re: (Score:1)
What is Intel? (Score:3)
Intel's self-image has long been that it is a semiconductor *manufacturing* company first and a semiconductor *design* company second. Going fabless would destroy the company's identity.
Intel has never been particularly good (Score:2)
Sounds like a horrible idea (Score:1)
Sure, why not. (Score:1)
Sure. Why should the US have any manufacturing capability at all? Especially if some executive can show better financials for two quarters in a row and get a beefy bonus.
Re: national security nono (Score:2)
A Trump victory would've shut down that IP-export ...
FTFY. The winning votes are in the mail.
Re: national security nono (Score:2)
Jebus where do you people come up with this stuff?!?
Honestly those of us who actually care about the national debt and lower taxes loathe the idea of a Democrat trifecta but at this point it is important to elect a sane human being , a group which neither you or Donald Trump belong.
Re: (Score:2)
Too bad there are no sane human beings running for President. Pence is the closest one available.
Re: makes sense (Score:4, Funny)
Intel is actually in Beaverton, Oregon. It is so boring in Beaverton that no one has the motivation to raise their voice, never mind riot.
Re: (Score:2)
Intel has many buildings they lease in Beaverton, but their fabs are in Hillsboro. It doesn't actually matter, because the cities are right across 185'th Avenue from each other, anyway.
Re: (Score:1)
Re: makes sense (Score:5, Funny)
God Hates Fabs
Re: (Score:1)
Re: (Score:1)