[HN Gopher] Intel doesn't know how to be a foundry, Tim Cook rep...
___________________________________________________________________
Intel doesn't know how to be a foundry, Tim Cook reportedly said in
2011
Author : retskrad
Score : 144 points
Date : 2025-01-29 17:53 UTC (5 hours ago)
(HTM) web link (www.tomshardware.com)
(TXT) w3m dump (www.tomshardware.com)
| raverbashing wrote:
| Well, he's not wrong
|
| Subdivisions that only work for one customer grow around all
| their idiosyncrasies and it's hard to adapt later
|
| Actually putting the design work into something you can work with
| is hard work
| duxup wrote:
| I've seen that with startups. I've seen that with established
| very successful companies that try to sell to new customers.
|
| You try to do something different, and every business process
| and management member is there to stop you because of what they
| learned previously ... when working with a different customer
| or market.
|
| The Steve Jobs story about resistance to developing a mouse in
| house is a good example.
| dlivingston wrote:
| What's the Steve Jobs mouse story?
| duxup wrote:
| This is the interview I'm thinking of:
|
| https://www.youtube.com/watch?v=s4Cz49MLh4o
|
| The general topic applies to what I mentioned but even just
| the specific mouse story is at this point in the video:
|
| https://youtu.be/s4Cz49MLh4o?t=90
|
| Folks who operate in a certain context / process or such
| and just can't imagine doing any differently and will stop
| you.
| zdragnar wrote:
| I've seen it kill more than one company. The company is so
| bent out of shape for that one special customer, that as soon
| as the relationship is over or isn't as profitable as they'd
| hoped, they aren't able to actually support any other
| customers.
| tyleo wrote:
| If you haven't listened to it already, Acquired remastered their
| TSMC episode. You can find that here
| https://www.acquired.fm/episodes/tsmc-remastered
|
| If you are interested in getting more context on TSMC, this is a
| great place to start.
| davepeck wrote:
| And to be clear, they also flew to Taiwan and interviewed
| Morris in English for what appears to be the first time in over
| 15 years:
|
| https://www.acquired.fm/episodes/tsmc-founder-morris-chang
|
| A tiny bit of this (excellent!) interview led to the story
| we're all commenting on.
| talldayo wrote:
| Sure they do. They just don't have any capacity for producing the
| chip density that Apple demands. It's like showing up to a
| motorcycle factory and complaining that they don't know how to
| manufacture cars.
|
| If America wanted EULV fabrication, it had to be organized and
| funded by the state. American fabs already made their business
| decision.
| CharlesW wrote:
| > _They just don 't have any capacity for producing the chip
| density that Apple demands._
|
| TFA says that the "Intel just does not know how to be a
| foundry" quote is from 2011 (14 years ago), and referred to
| Intel's lack of a "customer-centric mindset" rather than
| specific manufacturing capabilities.
| wtallis wrote:
| The article is talking about Intel's capabilities almost 15
| years ago. EUVL is not part of the conversation. This was when
| Intel was bringing up their first FinFET process.
|
| It's pretty widely known and documented that Intel at that time
| was in a horrible position to be a foundry for outside
| designers, especially ones that wanted to be able to design for
| more than one foundry.
| scrlk wrote:
| > If America wanted EULV fabrication, it had to be organized
| and funded by the state.
|
| The DoE funded initial research in to EUV via the national labs
| and EUV LLC back in the 90s. The IP was licenced to ASML,
| whilst Canon and Nikon (the leaders in lithography at the time)
| were blocked.
| Tostino wrote:
| People in government in the 90s made a good call, and we
| haven't done anything since then to ensure it actually played
| out, until the post WW2 world order was already extremely
| unstable.
|
| We have had multiple rounds of "why are we paying for any of
| this?" In our federal government since then.
| 01100011 wrote:
| In 2011. Seems like that should be in the title.
| behringer wrote:
| > "When the customer asks a lot of things, we have learned to
| respond to every request," Chang said. "Some of them were crazy,
| some of them were irrational, [but] we respond to each request
| courteously. [...] Intel has never done that, I knew a lot of
| customers of Intel's here in Taiwan and all [of them] wished that
| there were another supplier."
|
| Companies just don't get it, that customer service is almost
| always the most important aspect of the company. The customer
| will put up with a lot of bullshit if communication is rock
| solid.
|
| Handing off your customer service to agents that don't have
| reading comprehension, who don't have any authority, or who are
| completely non-understanding is going to hurt business.
|
| And I'm not talking about customer service that can write flully
| emails thanking and apologizing and butt kissing, I'm talking
| about good customer service is when the agent understands the
| request, takes it seriously, runs it up the flag pole if needed,
| and can act on it.
| duxup wrote:
| I have to manage my personal urge to reject or overreact with
| customers all the time.
|
| Customers do come up with a bonkers ideas / UI that's
| horrendous or just unworkable internal logic. And yet if I sit
| with them and we talk about "what can we do to solve the
| problem / enhance the product / get your goal done" I find they
| often come up with some good ideas and are happy to discard
| their initial madness.
| ChrisMarshallNY wrote:
| That's "Beans and Noses":
| https://archive.uie.com/brainsparks/2011/07/08/beans-and-
| nos...
| duxup wrote:
| That was great, thanks.
| georgeburdell wrote:
| (2011). And Cook was right. I worked at Intel for a few years
| during that decade and the foundry efforts were just not set up
| for success; in my area, they hired a bunch of new people, put up
| a firewall between us internal folks and the foundry folks, then
| without any guidance turned them loose. I was not even allowed to
| talk to them to troubleshoot equipment issues. They also got all
| of the equipment that we'd rejected for various reasons like poor
| process control, so they were newbies with worse equipment trying
| to start up a new group without help beyond what vendors would
| provide (for $$$)
|
| I have no insight into the customer facing side
| michaeljx wrote:
| interesting perspective. Care to elaborate a bit more? Where
| you in the design department? Why did they put up chinese
| walls? Was it to enable the foundry to have other customers
| other than Intel designs? Why did you have similar type of
| equipment? Were you also manufacturing chipsets? If so, why
| didn't they expand your division to become a foundry?
| scrlk wrote:
| There's a good video from Asianometry ( _Lessons from Intel
| 's First Foundry_):
| https://www.youtube.com/watch?v=-Y9LWYmVQu0
|
| If you're an Intel foundry customer, you don't want your
| design or IP leaking across to Intel's product team, who
| might be a competitor.
| petra wrote:
| But the IP is already "compiled", or if you want to isolate
| even further, separate just the mask making process. What
| could be learned from masks?
| B1FF_PSUVM wrote:
| > What could be learned from masks
|
| That's like asking what could you learn from building
| plans ...
| thetopher wrote:
| Sounds like they were following some advice from "The
| Innovator's Dilemma" by separating the business groups.
| nradov wrote:
| This wasn't something with the potential to become a
| disruptive innovation, as described in that book. Foundries
| are an established business model.
| georgeburdell wrote:
| I was in a fab module. The firewall part makes sense to not
| cross pollinate IP internally vs. externally, but it was
| taken to the extreme and management moved zero internal
| employees over to external so it was Intel's tools and
| recipes but not the talent who knows all the tribal stuff.
| ethbr1 wrote:
| > _but not the talent who knows all the tribal stuff_
|
| Tribal knowledge doesn't appear on management org charts or
| in human resource titles.
| twoodfin wrote:
| Given what we know about Apple and Cook, this seems like at least
| a modestly foolish story for Chang to have told.
|
| Apple's senior leadership has always for better or worse had long
| memories and held even longer grudges. And they don't like
| partners to speak for them, or create the impression--even if
| accurate--that Apple has a dependency on a single supplier.
| greedo wrote:
| That's true when Apple has alternative suppliers, ie graphic
| cards when nVidia blabs before a presentation.
|
| But who is going to replace TSMC? Sure someone will in 15-20
| years, but for now...
| Etheryte wrote:
| You could've said the same about Intel at one point, but here
| we are years later.
| johnnyanmac wrote:
| 2011 article, so GP's timeline fits.
| caycep wrote:
| they have used Samsung before...maybe with middling results
| but if push came to shove, Samsung's fabs have been the
| backup I think.
| mushufasa wrote:
| Chang is 93 years old. He can say whatever he wants.
| djcooley wrote:
| Second this. I don't think most people on hn understand how
| powerful he is.
| scarface_74 wrote:
| What is Apple going to do? Not use TSMC?
| alex-mohr wrote:
| Clearly the next step after building your own CPU and SOC is
| to start Apple Foundry and become totally vertically
| integrated?
| epistasis wrote:
| Particularly relevant with the 100% tariffs that Trump is going
| to try to ram through:
|
| https://www.tomshardware.com/tech-industry/semiconductors/ta...
| samtheprogram wrote:
| They're already running a plant in Arizona, and I heard it's
| going really well. It won't matter.
| wtallis wrote:
| The plant in Arizona is deliberately not equipped to produce
| current-gen chips.
| th0ma5 wrote:
| That is in an international economic zone and would be
| subject to tariffs as well as for supplies.
| WillPostForFood wrote:
| The point of the foreign trade zone is no tariffs, avoid
| customs. And the point of the tariff is to get plants built
| in the US, so tariffing them would make no sense.
| bryanlarsen wrote:
| Did you forget the sarcasm flag?
| duskwuff wrote:
| I wouldn't count on that. It's going to be some time before
| the Arizona plant is fully operational, and it's not a
| substitute for the production capacity in Taiwan. It's also
| very possible that TSMC will pause or halt production in the
| US if they're targeted by trade sanctions.
| patmorgan23 wrote:
| The TSMC Arizona plant is already operational they started
| production of TSMC 4nm process node this month, and will be
| ramping up volume over the next few months.
|
| https://www.reuters.com/technology/tsmc-begins-
| producing-4-n...
| GeekyBear wrote:
| TSMC is also looking to add a chip packaging plant in
| Arizona.
|
| https://pr.tsmc.com/english/news/3174
| TSUTiger wrote:
| Amkor will also be near
| duskwuff wrote:
| Even so, the ramp-up to full capacity is going to take a
| while - and even at full capacity it's not going to be a
| replacement for the facilities in Taiwan.
| jandrese wrote:
| They don't have to replace all of the Taiwanese
| manufacturing, just the chips that are US bound. There
| are plenty of customers in other countries that need TSMC
| chips.
| mandevil wrote:
| Besides the fact that the AZ plant does not produce TSMC's
| bleeding edge process (3nm) but the older 4nm, the product
| from that fab can't be delivered directly to a customer.
| Every wafer at that plant has to be shipped back to Taiwan
| for packaging and can only then be delivered to a customer.
|
| So a blanket tariff would definitely apply even to the
| production of that plant. Which is why lots of CEO's
| (including Tim Cook) are trying to suck-up to Trump right
| now, to get an exception made for their imports.
| indrora wrote:
| TSMC can work with partners for packaging (TI and IBM come
| to mind)
| GeekyBear wrote:
| Especially now that Taiwan has removed restrictions on
| producing TSMC's most advanced nodes on foreign soil.
|
| https://www.techpowerup.com/331104/tsmc-granted-
| government-p...
| scarface_74 wrote:
| Cook knows how to kiss up to Trump. There is a reason Apple
| wasn't affected by China tariffs the first time .
| ksec wrote:
| >TSMC founder says Tim Cook told him in 2011 that Intel did not
| know how to be a foundry
|
| That is the Original Title. It could be edited as "Tim Cook told
| TSMC in 2011 that Intel did not know how to be a foundry" - That
| would still have been accurate with the date on.
|
| Now the title has been editorialised, and meant or imply
| something else. Like most of the comments are already suggesting.
| insane_dreamer wrote:
| Just think where Intel would be today if they had truly leaned in
| and made the necessary changes to supply the iPhone chips.
| wmf wrote:
| Tim Cook would have negotiated away their profit margins?
| scarface_74 wrote:
| And how is that different than what is happening now?
|
| https://www.intc.com/news-events/press-
| releases/detail/1716/...
|
| > _Third-quarter GAAP earnings (loss) per share (EPS)
| attributable to Intel was $(3.88); non-GAAP EPS attributable
| to Intel was $(0.46)._
| xmodem wrote:
| Yeah, like they've done to TSMC over the past decade. (/s)
| foota wrote:
| Wouldn't they still be behind on the chip technology curve?
| Seems like this would hurt Apple, not dramatically help Intel.
| BeetleB wrote:
| This is back in 2011, when Intel was first flirting with the
| Foundry business, and has little bearing to their efforts today.
| That first effort mostly died. The goal wasn't to become a major
| player like Samsung/TSMC, but to ensure factories don't sit idle
| (i.e. just fill in the gaps with small customers). Intel products
| would get priority.
|
| The current vision is very different: It's to somewhat separate
| the fabs and Intel's products, and the end goal is that Intel
| Products will just be another Foundry customer.
|
| Whether they can achieve this is another story.
| jonathaneunice wrote:
| This situation has been and maybe always will be common. Those
| who view themselves as integrated manufacturers of *products*
| often cannot equally package, deliver, and sell their component
| *services*. No matter how good those services are at delivering
| the end product--and sometimes they are completely world-class or
| genuinely unique--they're almost always highly designed, refined,
| and evolved over years to map to the specific end products and
| their one business. They are NOT designed, refined, or evolved to
| be reusable, retargetable capabilities for all interested
| parties.
|
| By any rights, the major systems manufacturers of the 2000s--
| Cisco, Dell, HP, IBM, Sun Microsystems--should have had an
| enormous leg up in becoming the planetary scale, "red shift"
| cloud computing providers of today. Let's kick major operating
| environment vendors of that era in for good measure--Microsoft,
| Red Hat, VMware. Maybe even Oracle and SAP. They had the
| technical capabilities, the economies of scale, the vendor
| ecosystems, the customer relationships, the financial might.
| "Permission to win." Yet, only a few have made it. Microsoft has
| made the transition with Azure, but it's been a long hard pull.
| Maybe a few participation trophies for IBM and Oracle. But as a
| group, where are they, cloud-wise? Not too impressive that AWS
| and Google and others that had no classic "required assets" or
| "permission to win" have wiped them off the playing field or
| kicked them deep into the "others" bin.
|
| You can call that classic innovator's dilemma, and there's a lot
| to that. It's insanely, unfathomably hard to disrupt your own
| current successful business for new delivery models and
| aspirations. But there's something else as well: services and
| products are different beasts. They take different mindsets,
| expectations, business models, and metrics. All the best to Intel
| in becoming that new thing, a service provider. But it not just a
| side gig or slight extension. It's a radically different thing,
| and it's no wonder the highly successful product Intel of 2011
| wasn't also a good service provider.
| gsibble wrote:
| Counter-Point:
|
| GCP and AWS came out of product focused companies that
| effectively converted to providing the services they used
| internally.
| ethbr1 wrote:
| I'd never really thought about "How did hyperscale cloud
| providers become cloud providers?" And it's surprisingly
| distinct.
|
| AWS: Bezos API memo -> internal infra services that were
| external ready -> polishing and exposing those one-by-one
|
| GCP: Internal architectural/technical excellence -> new org
| that attempted to productize v2.0 of those services
|
| Azure: Believe it was mostly ground-up build as a new
| org/product?
|
| (No idea where Oracle cloud came from, internally)
| scarface_74 wrote:
| AWS was never an offshoot from Amazon.
|
| http://www.quora.com/How-and-why-did-Amazon-get-into-the-
| clo...
| LinuxAmbulance wrote:
| That's not true.
|
| AWS started from a one pager that two employees in the
| Capetown office wrote (forget their names, it's been
| years) about the potential to sell compute as a service
| from the newly revamped Amazon.com website backend and
| servers.
|
| Once the commercial viability was clear, Amazon.com
| created Amazon Web Services. The term 'offshoot' is
| perfectly reasonable.
|
| This is from working at AWS circa 2013, which was just a
| few years after it started.
| scarface_74 wrote:
| Again the link I posted was from Werner Vogels. The
| architecture that the CDO (Amazon Retail) uses is
| completely different than AWS.
|
| Have you seen the difference between the architecture of
| CDO and AWS internally?
| twunde wrote:
| It's rare but does happen. And frankly I'd only include AWS
| in the counterpoint. Google really struggled with GCP.
| Outside of Bigquery and Spanner many/most of the services
| were custom built for GCP and were not used internally. Hell
| they built a VM service when basically everything ran on Borg
| internally
| scarface_74 wrote:
| This is not true as far as AWS and this is a myth that needs
| to die.
|
| http://www.quora.com/How-and-why-did-Amazon-get-into-the-
| clo...
|
| AWS was always purpose built and designed as a new product
| and Amazon Retail has a completely different architecture
| from AWS.
|
| True, now some Amazon workloads have been moved over to AWS.
| But Amazon Retail is treated like an AWS customer.
|
| Even internally, AWS employees use a different system to
| stand up internal sandboxes than CDO (Consumer Division
| Operations).
| kridsdale1 wrote:
| Confirm that the majority of Google first party stack is
| not sitting on GCP. They are parallel.
| Maxatar wrote:
| The Quora link reinforces OP's point though, it doesn't
| contradict it at all.
|
| The only myth listed in your link is that Amazon made AWS
| to sell off excess computing capacity during periods of
| downtime.
|
| This is from the CTO of Amazon himself:
|
| >At Amazon we had developed unique software and services
| based on more than a decade of infrastructure work for the
| evolution of the Amazon E-Commerce Platform... The thinking
| then developed that offering Amazon's expertise in ultra-
| scalable system software as primitive infrastructure
| building blocks delivered through a services interface
| could trigger whole new world of innovation as developers
| no longer needed to focus on buying, building and
| maintaining infrastructure.
| scarface_74 wrote:
| Even today, if you look at the architecture and systems
| used by CDO and AWS, it's clear that AWS was designed
| from the ground up to be sold to customers. The
| architecture used by Amazon Retail was not suitable for
| third party customers.
|
| There is almost no similarity between AWS and CDO. I'm
| telling you this as someone who worked on the inside.
|
| At most they took some of the things they learned. But
| it's not like they took the code from any Amazon Retail
| implementation and did the equivalent of a "git checkout
| -b aws"
|
| S3, SQS and EC2 the early AWS services were not based on
| the Amazon Retail systems at the time.
|
| The entire control plane of CDO is completely different
| than AWS and always has been
|
| If you want to see what services look like thst started
| out on Amazon Retail and moved over to AWS with few
| modifications - look no further than Amazon Connect. It
| was a lift and shift of the call center software that
| Amazon Retail uses without any AWS integrations or even
| publicly available APIs at first. It's gotten better
| since 2020 when I was at AWS.
| Maxatar wrote:
| But none of those claims contradict the point that was
| made. It's not really clear what you're disputing.
|
| AWS exists because Amazon had already developed the know-
| how and expertise in how to deliver computing
| infrastructure as a service and saw that offering this as
| a service to third-parties would be profitable. The only
| myth is that Amazon wanted to sell off excess computing.
| scarface_74 wrote:
| There is a huge difference between "we have the know how
| to create AWS" and "we used spare capacity to sell to
| customers" or "we took code we used from Amazon Retail
| and created AWS".
|
| Most people think the former and even now people seem to
| think that there is any similarity between how Amazon
| Retail (CDO) and how AWS runs. Again with the caveat that
| some of Amazon Retail's workloads have been migrated to
| AWS or were born on AWS.
| Maxatar wrote:
| Exactly, there is a huge difference, that's the point!
|
| You seem to think the original claim is that someone took
| code from CDO and built AWS off of it, and no one said
| that. The simple claim which is reinforced in the very
| Quora link you provided is that Amazon had developed
| expertise in delivering computing infrastructure
| internally, and decided it would be profitable to offer
| this service externally.
|
| Your claim that most people have a very specific belief
| about an implementation detail regarding whether literal
| code was reused or transferred is almost certainly false.
| The belief is that a company that developed expertise in
| an area that proved to be invaluable internally managed
| to leverage that expertise into a new product line that
| they could sell to other people, and nothing you have
| provided contradicts that.
| scarface_74 wrote:
| This is the original message
|
| "GCP and AWS came out of product focused companies that
| effectively converted to providing the services they used
| internally."
|
| They at most used thier know how. You and I are in
| agreement. But that's not what is implied by the parent.
| ghaff wrote:
| AWS was obviously not going into building datacenters and
| taking a modern service oriented architecture approach
| cold. But, as you say, the pervasive myth is that AWS
| started out by using excess retail computing capacity and
| presumably the same architecture which, by all accounts,
| is simply untrue.
|
| And GCP more or less started out as a classic PaaS. Azure
| really had more of an on-prem focus at first.
| alex-mohr wrote:
| Also a myth for GCE.
|
| From a technical perspective, App Engine and Compute Engine
| were built on top of internal infrastructure (borg), but did
| not expose borg directly. And there were a number of
| interesting mismatches between the semantics that customers
| expected of VMs and what borg offered to its containers that
| eventually resulted in dedicated borg clusters with different
| configs for cloud. And some retrospectives on whether
| building on borg was a better option than going bare metal
| directly.
|
| Org-wise, the App Engine team was first and not part of the
| internal-focused Technical Infrastructure teams. GCS came
| next, and it too was not part of the canonical storage org.
| Then GCE, which was only possible because it was either
| written off or at least tolerated as an experiment by most,
| with a few key people providing behind-the-scenes support to
| make it happen -- especially in networking. It likely also
| helped that GAE was in SF and the rest of GCP in
| Seattle/Kirkland initially, so geo provided some insulation
| too.
|
| The dominant perspective internally was that Google's
| technical infrastructure was its secret sauce, so why would
| they give it away to others? It took a long time to change
| that.
|
| [Disclosure/source: I was on GCE and helped get it launched.]
| npalli wrote:
| Less of a technical reason than the fact that TSMC was willing to
| go farther in accommodating Apple's demand (which is right).
| Also, this is Morris Chang relaying, so not totally unbiased.
| The implication was that Intel lacked the customer-centric
| mindset required for a foundry business. Unlike TSMC, which
| tailors its process technologies to meet customer needs, Intel
| was used to designing and producing its own chips and struggled
| to adapt to servicing external clients. By contrast, Apple valued
| TSMC's ability to listen and respond to specific demands,
| something Intel historically did not do.
| timewizard wrote:
| "Intel makes more than enough money that they don't have to
| cater any part of their design to external partners and
| historically have done very little of this type of work. Our
| use case was simply not on their radar."
|
| On the other hand I can appreciate that Intel chips are more of
| a known quantity than the various ARM designs that are floating
| around out there.
| scarface_74 wrote:
| And just a note, Intel is losing money and still as of the third
| quarter of last year paying a dividend while complaining about
| cash flow
| xyst wrote:
| Thanks to neoliberal economic theory, the public CEOs must
| deliver shareholder value at all costs! In this lens, they are
| doing "great"
|
| But actual company itself and products are bleeding internally
| from decades of mismanagement.
| BeetleB wrote:
| Intel suspended the dividend.
| kmeisthax wrote:
| So, Intel's fault was being focused on selling you the whole
| widget and telling you what you wanted, instead of something that
| met your needs. Just like Apple, IMO - the whole "product vision"
| thing is great when it's about figuring out how to sell a good
| smartphone and terrible when it's selling iPads with desktop
| chips in them that can't do any useful professional work outside
| of drawing.
___________________________________________________________________
(page generated 2025-01-29 23:01 UTC)