Intel in talks to buy GlobalFoundries for about $30B
reuters.com
reuters.com
From Intel's perspective, there are also some US assets from IBM, which GloFo "acquired" in 2014 (for -1.5 billion at the time), which includes some oldish US fabs, but also a patent portfolio. It certainly goes into the current US political narrative that Gelsinger has been pushing.
But perhaps more importantly, I think that what Intel is buying is what was sorely lacking them : people and a methodology of working with external customers. From all I've heard over the years, Intel's process and design teams are so insular and out of touch with the rest of the industry's practices that they had a terrible time just working with customers. It wasn't pretty when they tried the two previous time to open up their process. Loads of announcements, few products if any coming out, and when it worked, they ended up buying their customers anyway, see Altera.
If Intel truly wants to go in the foundry direction, they'll get some interesting outside experience, and while it may not make sense from a pure process point of view, considering how much of a sore point their inability to make it work has been on their previous attempts, the idea may have a lot of merit.
Perhaps a very smart move from Gelsinger, time will tell, in the meantime I'll have a laugh at Intel buying AMD's old fabs :)
You used the term "acquired" along with a minus sign in front of 1.5 billion but just in case other people don't get the joke, IBM gave their fabs to Global Foundries and PAID them $1.5 billion to take them.
Yep. I'm glad others are aware of this. There were certain manufacturing processes that needed to be sustained for "national security" reasons. GF was an imperfect solution but it was better than the alternatives at the time.
Anyway I wonder if GlobalFoundries will be it's own business unit or otherwise get enough independence to do things the right way / differently than in the past? I often find large companies who are 'not good with customers' aren't really great at acquiring competency at being good with customers if they just buy and assimilate it.
Heck I've been a part of companies who had a big customer focus and they acquired smaller companies who weren't good with customers, and somehow the smaller company infected the organization and suddenly they became bad with customers :(
However, if they did buy GloFo it would give them a straight forward path to having a "Fab for hire" service ala TSMC and their own fabs as well. Kind of a have your cake and eat it too thing if you can bring GloFo's ability to bake chips up to the requisite standard.
This is such a big deal that it presents problems for vendors working on even completely open, non-sensitive systems. Like my company. We are a foreign language instruction company that serves primarily DoD. Some of our customers are being super reluctant to approve a specific project of ours because it is nearly impossible to buy appropriate hardware that is not manufactured in China. We have to settle for a hardware vendor that is not our favorite choice because at least the design company is US owned.
It's not technically correct. It should be fine to use these systems for non-sensitive purposes, but you try negotiating a contract with a CO who won't explain themselves/refuses to clarify with his own command that this is OK.
I am a bit interested in the relationship with Abu Dhabi.
Of course not all prior art is in patents so any journal publication could also force narrowing of claims. Older patents are almost always broader... if they cover the thing you want at all.
Turns out everyone who does X also needs to do A, to the extent X is X+A, so everyone needs to license both patents.
Then, I can patent a cart with “a number of wheels” and “at least one handle”.
That’s why patents tend to be so vague.
You don’t want to keep the tiniest door open for such cases, so you don’t use a mirror, you use a light-reflecting device. If you’re smart, you don’t even use the term ‘light’. Your invention likely works with radar and röntgen, too. Maybe avoid electromagnetic fields, too. Who knows whether something similar, in the abstract, is possible with sound?
You also don’t say a device needs to do A, then B if there’s a tiny chance it might be possible to accomplish the same doing B first, then A.
But yea generally you want to describe things with a billion examples, but no statements limiting the scope of your invention.
The most devious strategy is that you disclose vague and probably impossible embodiment. Wait 10 years and then write claims that that narrowly target big products that you really couldn't have invented ten years. But the burden is more or less on the accused infringer to provide you didn't describe the new invention in the old patent.
This is not a problem at all. If they wanted to go this direction it's as simple as dumbing down, cheapening their process a bit, and giving free reign to make decisions on what they will be fabbing to technically competent salespeople.
With Intel's resources they could've even developed such commodity process separately to their own process, and ran them in parallel.
They had almost a whole decade to do that. It's just they simply didn't.
They've wanted to go in this direction for many years and so far failed. We've had meetings with them over the years and both times their offering was uncompetitive. I'm not sure what the sales people are supposed to do if the core technologies fabless customers want are not available.
I lot of people don't realise Intel are already in the foundary business. I'm not surprised their customers are not shouting about it - why would you want to tell everyone you are using an arguably failed process.
Can someone with insight explain how this can happen? GloFo is a company worth billions, how can they just fail at this while Samsung succeeds. What's so hard about this in particular? I don't think I recall companies at this scale just failing at something that is supposed to be their area of expertise.
As to what failing means, it can be many things, sometimes you just have way too many defects to have the process be commercially viable (if your cost is 10x your competition and you can't fix it, you don't have a process), some bad technical choices because there's been less and less cross industry cooperation (things like the ITRS that gave a rough roadmap hasn't been a thing for a while, last I checked), and introducing new technologies (EUV or new metals) may compound those. Sometimes chips in your new process are more dense (which is good), but perform poorly because of one of the previously mentioned stuff to the point it's hard selling new chips that perform worse than previous gen. In Intel's 10nm case there was a bit of everything and it will be very interesting to know exactly what factored in more (Cobalt is often rumoured, a very overconfident choice for MPP was my initial understanding, I haven't really followed on this).
In the case of GloFo, it's perhaps simpler than that : GloFo was part since 2007 of an alliance between them (initially as AMD), IBM and Samsung called the Common Platform [1] that had more or less standardised a process, with a lot of the heavy research done by IBM (I would say, out of spite for Intel) and then used by GloFo and Samsung.
14nm was the first that everyone was on their own, because IBM failed their research. In the end, Samsung managed to get it done, and GloFo either failed or... didn't try really hard ? This is where I'm gonna speculate a tiny bit and be less charitable, I don't think that GloFo was sufficiently funded in terms of process development, despite their parent owner being able to.
And it definitely looked like from the outside, at least from my point of view, they were always in dire need of money except for acquisitions, and late with everything (including the process inherited from the alliance). Most of the business decisions done (including buying Chartered and reverse buying IBM's 22nm fabs) seemed to be about pure business and missed the technical understanding of the business and what it required to keep it going.
At the end of the day, for them, licensing 14 from Samsung was a great out of that mess. But that didn't happen for 10, and we (at least I) don't really know why. My guess is that GloFo would probably have been happy to keep going on with that arrangement, but I may be completely wrong on that one.
At the time I seem to recall that both had some issues filling up their order book (which seems a bit insane today, considering the current shortage) and that may have played into it from Samsung's point of view.
[1](https://www.chiphistory.org/546-ibm-microelectronics-common-...)
For whatever it's worth, it feels like GloFo has had various process issues for a -long- time. While there were issues with the Bulldozer design itself from a deep pipeline perspective etc, the other factor in it's lukewarm lifetime (especially the first couple gens) was issues with GloFo's processes even back then. I know some of that was that they were trying to gear more towards bulk silicon vs CPUs but I think there were other issues too.
I think no process could have saved Bulldozer for AMD, it truly was a deeply flawed design IMO, but yep GloFo certainly didn't help. I think my "uncharitable" take was a bit too kind considering.
Specifically in the case of AMD, GloFo relied early on on guaranteed income from AMD through the Wafer Supply Agreement which was stupidly tight (my understanding is that it was part of the deal of selling to Mudabala, though exact details are not public), included a massive volume of must buy from AMD, and didn't include any penalties for GloFo missing deadlines or having their processes underperforming. They definitely coasted on this and I should probably have mentioned this above, since it led to many of their issues going forward till 14nm.
It took them years to manage to loosen the WSA up, the last step was (if I recall correctly ?) loosening the CPU core exclusive that I believe was part of it (they had a carve out for GPUs which were built at TSMC at the time of the sale, though that was supposed to be phased out if my memory is correct, before it was renegociated and AMD moved the low end GPUs there instead, except maybe one gen?).
To their credit, AMD pretty much never bad mouthed GloFo in public. Managing to do the original Ryzen at GloFo and getting as much out of the 14 process that they did is, I think, one of the most underrated engineering tour de force from those who worked on that. While supposedly "copied exact" from Samsung, it wasn't exactly fully there at the time.
I did read that the last renegotiated WSA between AMD and GloFo is supposed to be going on till 2024 (?) currently, so if this goes through, I hope there's a "in case of sale" carve out, if just so anyone can save face a bit ;)
(and that would be a pretty large unburden from AMD in any case, which makes the whole thing even more ironic)
Only the IOD are still on GF 14 / 12 nm, along with some APU. Everything else are TSMC.
This Intel situation is certainly part of "keeping it in the USA" though it would have been far easier and cheaper to prevent outsourcing back in the mid-1990s when Silicon Valley's silicon ebbed out of the country. It was a night-and-day change and lots of knowledgeable people were scattered to the wind back then. Now much of that knowledge is having to be reinvented and rediscovered from scratch (because it was tacit knowledge that never could be recorded into books or other writing).
This probably is the best way for Intel to become a foundry. They tried 10 years ago and the entire process was 100% Epic Fail. Not surprising given the pain seen when AMD split off its fabs to form GF, when it abandoned being an IDM vertical fab. When you are an IDM, many process-design interfaces break or have to be formalized in ways that were never imagined or learned by the organization. That's how Intel failed its foundry attempt before so simply avoid Intel culture by buying an already working culture makes a lot of sense.
You don't just decide to make a chip. There are hundreds of foundry specific libraries, design rules, and more. There are always problems with the EDA CAD tools for new processes and you need lots of support people to answer questions and help debug problems that your customers are having.
Intel has tried being a foundry before multiple times over the last 20 years. They always fail and give up after a couple of years.
Considering the way things are going I am more than a bit intrigued, but not surprised.
and then Aliexpress buys Amazon.
But the chance to utilise EMIB would be cool.
That would be the concern. What happens to the chip market when intel rediscovers its inherent anti-competitive nature and shuts down any non-intel product?
I didn't realize they still had a sizable US presence. Malta NY (14nm fab) & outside of Burlington VT (90nm fab). The largest employer in VT! Also Singapore & Dresden (12nm).
Supposedly they were going to try to go public. To be honest I have a hard time understanding what's in this purchase for Intel, hard to see this as anything but anti-competitive, but Intel is trying to become a silicon foundry to others & GF knows how to do that. They also have a decent patent warchest.
If the argument is facilities and staff, that's fine. But buying an outdated fab that won't compete with TSM isn't that big of a game changer, Intel already has tons of that.
https://www.anandtech.com/show/16621/tsmc-q1-2021-process-no...
Even Nvidia's RTX 3080TI is on an optimized version of Samsung's 10 nm node.
While semiconductor fabs don't have the useful life of say a railroad, they are valuable for far longer than the 2-3 years they are in the spotlight. I'm no expert in this area, but as the recent semi shortage demonstrates, it appears today's 'ancient' fab can be viable for at least 20 years if not longer.
Given the astronomical costs of both the fab and chip design for current nodes, we're likely to be seeing <=28nm sticking around for far longer to get their payback on the process nodes. So if you're Intel and you gobble up as much of the production capacity as possible, that would seem to be a prudent move as long as you don't overpay relative to how much you'll be able to increase prices with the greater share of the market. Think like Intel probably is: they're not thinking of being one of many semi players, they're thinking of being one of the last 2-3.
1) No money has been received. USICA ($52b number) has not even been passed through congress yet.
2) 12nm is hardly “outdated”, not every use case requires the most cutting edge node. There is plenty of value to be had at that process size.
3) There is more to running a foundry than process size, and Intel is likely interested as much/more in the customer relationship parts, as elaborated extensively in other threads
What are they supposed to do, not compete during the heyday of their industry because they expect a tax credit to be paid out over the next several years?
>Sinema’s bipartisan CHIPS For America Act - which is now law - restores America’s leadership in semiconductor manufacturing.
https://www.sinema.senate.gov/fueled-sinemas-chips-act-intel...
https://www.congress.gov/bill/117th-congress/senate-bill/126...
The CHIPS Act is not USICA. That $52B number comes from USICA, which has not yet passed the House. I was being deferential earlier out of politeness but you didn't take the hint and look it up. Here is some light reading on it:
https://www.engadget.com/senate-usica-endless-frontier-act-0...
https://www.theverge.com/2021/6/8/22457293/semiconductor-chi...
And my comment specifically referred to the chips act, which has passed [1]. I'm not really sure why you would be distracting from the point by bringing up other legislation that has not yet passed? Like all good corruption, exactly how much money is moving around is obfuscated. I've seen both "$10-$50 billion" and "$52 billion" as numbers for the CHIPS Act. If you have a more accurate number I'd be more than happy to update my post.
So the question is did they get billions from the chips act like the senator that wrote it thinks they did or not? If they did then you're quibbling over exactly how many billions they got, which isn't a good look. That they potentially have more coming in doesn't really change the substance of my comment in the context of the original article at all. So thanks for the heads up on the next few billions they will be getting I guess?
[1] https://www.sinema.senate.gov/congress-approves-sinemas-bipa...
Intel has historically been willing to play dirty, like messing with compilers to benefit Intel over Amd[2]. Or like when they gave kickbacks to OEMS [2]. Or when they paid vendors to not support AMD fully in software[2]. Or just plain old false advertising [2]. Actions like these is what led to a 1.25 Billion dollar settlement with AMD [3].
Outside of playing dirty, it is a good strategic move for their new foundry business. You get the tools, know how, processes, and most importantly customers that can make the venture a success. Intel has had a foundry business before [4], but they closed it. "ecosystem is everything with the foundry business and that takes time, money, and technical intimacy," [4]. Intel has money, and they are buying the years in the foundry business they missed along with technical intimacy to more standard tools. So, I think it makes sense from a business perspective, as it gives Intel a recurring revenue for its older processes and in turn greater advantage from its vertical integration.
[1] https://www.anandtech.com/show/15038/globalfoundries-and-tsm... [2]https://www.anandtech.com/show/3839/intel-settles-with-the-f... [3] https://www.cnet.com/news/intel-to-pay-amd-1-25-billion-in-a... [4] https://semiwiki.com/semiconductor-manufacturers/intel/7912-...
OTOH my impression was that GloFo stopped short of EUV due to the cost, not that they couldn't do it. Intel might be able to fix that, but I think they're probably going the way of IBM.
Allowing smaller market participants to consolidate to face the behemoth is not necessarily going to be perceived as anti-competitive.
Intel is not small by any measure. Now TI and GloFlo, or some other pairing I might agree with.
1: https://www.sinema.senate.gov/fueled-sinemas-chips-act-intel...
At least that’s typically how these things go. I hope it’s different ‘this’ time.
(As far as I know, AMD always held long contracts, partially with exclusivity, with GF for fabrication, although this has weakened due GF falling behind and not meeting its promised process targets)
My understading is that AMD only used said contract for the I/O part of their chips or some entry level GPUs, as they moved the CPU cores to TMSC. I'm unclear on their current APU lineup.
https://videocardz.com/newz/amd-monet-12nm-zen3-apu-and-berg...
> AMD will see out its existing commitment to use GlobalFoundries through 2024, with the latest amendment setting purchase targets for 2022, 2023, and 2024. Beyond those new targets, however, the agreement releases AMD from all further exclusivity commitments to GlobalFoundries. AMD is now free to use any fab on any process node that it wants.
1.6B is half a quarter's revenue for AMD. Milan has been a slam dunk; there's obviously some value over at GloFo for AMD, and it wouldn't be a no-tears breakup (hypothetically speaking).
I don't dispute that GloFo's role in AMD's strategy will necessarily change if it becomes part of Intel.
(a) Intel is a distant third in fabrication and a merger with another also-ran would improve competitiveness against state-backed TSMC and the Samsung chaebol;
(b) the severity of the demand creation in the market justifies an acquisition that streamlines existing players and makes them more productive than the sum of the deal's parts; and
(c) bringing GloFo under a US-owned corporate umbrella is vastly beneficial to the national interest in a way that only Intel can fully enable.
It'll be an interesting move if it comes through to fruition. TBH this would ironically probably be a good thing for Taiwan and may be a good thing for TSMC.
It is somewhat amusing that they're picking up AMD's scraps.
Intel was pretty aggressively driving down price-per-core up until the 6000 series/ 14nm chips. For example here's the price series for a 6-core chip from each of their series:
i7 970 (Nehalem-E, July 2010): $885
i7 3930K (Sandy Bridge-E, Nov 2011): $583
i7 5820K (Haswell-E, Aug 2014): $389
i7 6800K (Broadwell-E, May 2016): $434
So in basically 4 years (+1 month) they reduced the price of their entry-level 6-core processors by more than half (56% reduction), comparing 970 vs 5820K. It specifically is Broadwell (the first 14nm chip) where intel lost the plot and prices started increasing.
People also dump on Intel for not increasing core count in the consumer platform chips, which I think is a deeper subject than people treat it as. Certainly by the 6000 series they should have pushed the consumer platform up to 6 cores, but fundamentally the consumer socket is driven by the economics of office machines and walmart PCs. Up through Haswell, 4 cores was the right amount for an average desktop for the average office worker (not talking about powerful workstations for developers here, those users always had the option of the HEDT platform with more cores/memory) or the average home user (again, not talking about power gamers, and if you wanted more HEDT was always a fairly reasonably-priced option too, especially during the Haswell generation). There was, at the time, no 6-core die, the next step upwards was the 8-core HEDT die with quad-channel RAM and all that other stuff, and I doubt that would have fit into a consumer socket even if you wanted it to, and having 6-core chips with dual-channel RAM probably would have impacted performance until DDR4 came along with the 6000 series.
On the whole Intel didn't actually do too bad up until the 6000 series, when the combination of a complete shitshow of a 14nm node (in these latter years of 14++++ people forget what a mess early 14nm was), the lack of competition, and the beancounters gaining firm control of the company all combined. But that's really a specific, narrow range of time - 2016 and 2017 mostly. Then in 2017 Intel shit their pants after seeing Ryzen and released the 8700K, which was a wildly aggressive product for Intel at the time - same i7 pricing but 50% more cores, and such a solid product that it basically forced AMD into similar price reductions (abandoning the $500+ segment for a generation, and big price cuts on their existing products).
I really see the "intel got greedy" thing as being specifically 2016-2017. 5820K was extremely good value for a 6-core (even if it did need a "more expensive motherboard", those are cheaper than a high-end X570 now and far far cheaper than sTRX40), the problem was the 6000 series and failing to push core count up on the consumer socket at that time, and I think that is largely attributable to early 14nm woes. Certainly by the 7000 series yields would have improved and Intel could have pushed core counts up, 7000 series is arguably one of the worst releases Intel has done in terms of how rapidly it aged out of the market.
(although to be fair - up until like 2018/2019 you could easily sell a 7700K and get almost all your money back and do a straight upgrade to a 8700K, so it wasn't the end of the world from a user perspective.)
yeah there were still ridiculously priced processors (funny how the "$1000 processor" memes got forgotten when 3950X came around) but there were also serious dealz on higher core count processors, if you made the effort to buy them. The consumer platform didn't hop right up to hexacore+, but they were there on the HEDT platform if you wanted them.
but yea 6000 and 7000 series they were milking it.
I can’t say either way for sure, I’m not an accountant at Intel. But I don’t think anyone else here is either.
AMD is re-surging with great technology, they already chipped away some market share in key sectors and I see that continuing for a while.
TSMC has closed the gap and exceeded Intel in manufacturing, Samsung is not too far behind. Intel is no longer the market leader in chip fabrication.
Nvidia is buying ARM and aiming to start competing in the CPU space, with their GPU leadership that could be huge.
x86 is starting to lose it's complete dominance & new players coming in(Apple M1, RISC-V, Nuvia, Ampere)
Cloud vendors are starting to build their own chips and tech instead of buying them(Graviton, TPU)
I see Intel still being profitable and stable but losing their truly unprecedented grasp on the industry.
I do agree they won't have "truly unprecedented grasp on the industry", because I think they lost that a long time ago with globalization. The existence of TSMC/Samsung etc. I do think that means simply a larger market with more players and customers, rather than a shrinking Intel.
Microsoft evolved their business with the times and came out stronger, I think Intel will as well. I know they've been in talks with Nvidia to manufacture some of their stuff due to offshore supply being uncertain at best. If the world continues to be as uncertain as it has been, I think they're in a very different position than the widely held popular notions that you described.
It was all good, even preferable, to outsource that dirty manufacturing in the good times. I believe all the commonly held analysis that Intel is set for decline no longer has legs to stand on. I don't view fabless design firms like NV/AMD/Apple/ARM as competition. In these dark days, if you can't make things, you basically have nothing. Just a piece of paper.
Intel still has a long way to fall. Unless ARM becomes dominant to the level x86 is I think they'll recover.
https://electronics360.globalspec.com/article/16465/report-t...
This has always been national-security leverage for Taiwan - if China ever attacked, Taiwan would make sure TSMC was destroyed and utterly cripple worldwide chipmaking. Like if you think the current shortages are bad imagine the only cutting-edge fab in the world disappearing overnight. That leverage only works if all of TSMC's facilities are in Taiwan so the idea that they would ever agree to build one anywhere else was absolutely unthinkable even a year ago.
The US must be playing some serious hardball given the shortages that have sprung up this year and it looks like they're getting results. First it was a little fab for national-security stuff and it's morphed into the US getting a gigafab.
On top of that Japan is getting a TSMC specialty fab - not power stuff, but it'll do CMOS sensors for cameras and automotive/embedded type stuff, and the EU is negotiating for a cutting-edge fab of their own. It's a huge sea change in the last year.
Long $TSMC i guess
That is actually the consensus reason (not sure it was ever stated outright) for the original plan for the 20k wf/m fab TSMC was building in Arizona (that has since been upgraded to 100k). Even if it wasn't necessarily a huge plant in terms of global supply, it gave the US defense/intelligence community a guaranteed supply of chips with domestic manufacturing and chain of custody.
It is nuts just how much of cutting-edge production is concentrated in the pacific rim region there, and the current shortages have really driven that point home even harder. Even leaving aside Chinese invasion of Taiwan, there could be a tsunami or some other natural disaster that takes out TSMC and that would just absolutely gut the worldwide economy. Imagine the current shortages where everything is still running at capacity, and then imagine that TSMC or some other major fab would go offline (Samsung getting hit in Korea, etc), be destroyed and take years to come back online - there are the same limitations as bringing new capacity online, you have to wait years to source new machines from ASML and so on.
It potentially could happen anytime. It could happen tomorrow. Just another big earthquake that sends a tsunami at just the wrong place and compounds all the current shortages even worse. Even apart from the national defense aspect, it is a dangerous concentration of irreplaceable supply chain elements and I think the US, EU, etc are really pushing to mitigate that danger a bit.
(No, really. See https://en.wikipedia.org/wiki/List_of_semiconductor_fabricat... )
What other companies also do this? Are Nvidia, and AMD on this list?
No and no. IBM (for Power products) and arguably Samsung (for their mobile chips) are effectively the only ones left on their own leading edge nodes. Everyone else uses TSMC, GloFo, or SMIC.
Big boys like Maxim, Marvel, Broadcom, Qualcomm, nVidia, AMD, Ti[1] don't have their own fabs.
What about memory? Samsung, Hynix, Micron all do their own stuff.
Analog ICs? National, Ti[1], NXP, Skyworks, ONsemi, ST, to name a few.
[1] Ti got rid of their fabs years ago to switch to TSMC but over the years they acquired a few companies small than them that have their own fabs. Those divisions still use them for fabbing those products.
National was acquired by TI ~8yr ago.
[Gestures at IE vs Netscape]
[Gestures at Chromebooks which 99% of population never run Edge or Firefox on]
(Yes, "the fight" is between Google and Apple now, but still...)
http://www.businesskorea.co.kr/news/articleView.html?idxno=6...
And if we look at the entire chip-making business, instead of the narrow foundry-for-hire category? Or consider the possibility of Intel expanding (not just buying) into the foundry business?
Also, note that anti-trust law isn't about the possibility of future expansion. Intel already tried to enter the non-memory foundry-for-hire business and failed.
Mind you, I've been at 2 organizations that have purchased chips from foundries, and we turned down both Intel and IBM's offers. I'd love to see more competition in that industry, and Intel buying GloFo will add to competition on the high end.
How far behind is Intel really?
Genuinely curious what you think.
GF won't have much if anything that helps them in that regard, so I think it mostly means that Intel thinks they can use more capacity of their own, even for not top-end process.