Intel Is In Talks To Buy GlobalFoundries For About $30 Billion (reuters.com) 57
New submitter labloke11 shares a report from The Wall Street Journal: Intel is exploring a deal to buy GlobalFoundries (source paywalled; alternative source), according to people familiar with the matter, in a move that would turbocharge the semiconductor giant's plans to make more chips for other tech companies and rate as its largest acquisition ever. A deal could value GlobalFoundries at around $30 billion, the people said. It isn't guaranteed one will come together, and GlobalFoundries could proceed with a planned initial public offering. GlobalFoundries is owned by Mubadala Investment Co., an investment arm of the Abu Dhabi government, but based in the U.S. Any talks don't appear to include GlobalFoundries itself as a spokeswoman for the company said it isn't in discussions with Intel.
Intel's new Chief Executive, Pat Gelsinger, in March said the company would launch a major push to become a chip manufacturer for others, a market dominated by Taiwan Semiconductor Manufacturing Co. Intel, with a market value of around $225 billion, this year pledged more than $20 billion in investments to expand chip-making facilities in the U.S. and Mr. Gelsinger has said more commitments domestically and abroad are in the works.
Intel's new Chief Executive, Pat Gelsinger, in March said the company would launch a major push to become a chip manufacturer for others, a market dominated by Taiwan Semiconductor Manufacturing Co. Intel, with a market value of around $225 billion, this year pledged more than $20 billion in investments to expand chip-making facilities in the U.S. and Mr. Gelsinger has said more commitments domestically and abroad are in the works.
An admission of failure (Score:4, Insightful)
That CFO turned CEO did tremendous damage to Intel.
Glad to see the new oldtimer setting things straight.
Sweet Irony... (Score:5, Interesting)
Re: (Score:2)
Yes, and more importantly the I/O chip for their EPYC CPUs are made there as well.
Re:Sweet Irony... (Score:4, Insightful)
Yes, and more importantly the I/O chip for their EPYC CPUs are made there as well.
It's a small world, and there's a lot more shared between the companies than most people assume given the competition. It's not like they've never worked together. The i7-8809G for example has AMD graphics and is more powerful in that regard than AMD's competing APUs at the time.
Intel isn't going to suddenly fuck over AMD if they buy glofo. It just doesn't work like that.
Re:Sweet Irony... (Score:4, Insightful)
Intel isn't going to suddenly fuck over AMD if they buy glofo. It just doesn't work like that.
Intel has a history of fucking over AMD both suddenly and gradually. The only question is which of those it's going to be.
Re: (Score:2)
Which? You are very optimistic.
Re: (Score:2)
Okay, which first.
Re: (Score:2)
GF 12LPP is pretty much Samsung 14LP with the serial numbers filed off. Sometimes 3rd party hard IP isn't available in both process, but for case of switching foundries this is as easy as it gets.
Re: Sweet Irony... (Score:2)
Indeed it was. Fascinating turn of events.
What goes around . . (Score:5, Funny)
Re: (Score:3)
I wouldn't make too much fun of them because AMD bought Xilinx [nextplatform.com] as well. And yes they have plans [hothardware.com] for it.
Re: (Score:3, Interesting)
AMD didn't do terribly well - this was after the Athlon era, and AMD's chips were relative flops. Divesting themselves of their fabs was seen as an emergency move to raise capital quickly to keep them in business.
In recent years AMD got lucky with Zen and is in a much better position now than they ever were. It was highly believed that it was Intel who got Microsoft an
Re: What goes around . . (Score:2)
That doesn't make sense, Intel wouldn't tell them to buy from a competitor, especially when they're doing particularly bad right now. It makes more sense that they were simply looking for a more powerful APU, and AMD has had the best for quite a while now.
Re: (Score:2)
Re: (Score:2)
It's not just Zen, it's everything... chiplets, 7/5nm, Navi, hypertransport, architcture that resists side channel attacks better, it goes on and on.
Re: (Score:2)
Who knows? Analog Devices recently gobbled up Maxim, so now we have National Semi, Analog Devices, and TI as juggernauts
Re: (Score:1)
Is Intel buying fabs or customers (Score:2)
Bad for everyone else. (Score:5, Insightful)
This might be a smart move for Intel but this is very bad for everyone else, especially if we want a competitive market. Mergers and acquisitions are not ways to promote research and development.
Re:Bad for everyone else. (Score:5, Insightful)
Mergers and acquisitions are not ways to promote research and development.
It's not about "research and development". They gotta keep all that fed money moving around. What better way than to "diversify" their portfolio for "Maximum Shareholder Value"®?
Re: Bad for everyone else. (Score:4, Interesting)
Depends (Android 1.0) (Score:3)
> Mergers and acquisitions are not ways to promote research and development.
That may be true in THIS specific case because there are very few players, and they are very large. Global Founderies can do R&D without Intel, and vice versa.
The general statement that "acquisitions are not ways to promote research and development", I would disagree. Often an acquisition is a large, well-financed company buying a company that has a cool idea and little else. Android Inc was facing eviction when they got a ma
Re:Depends (Android 1.0) (Score:5, Informative)
There were good reasons [forbes.com] for that and it looks like it's all paying off. [anandtech.com]
Re: (Score:2)
Re: Bad for everyone else. (Score:2)
TSMC and Samsung are still around.
Re: (Score:2)
But for how long?
Civil War (Score:5, Interesting)
Re:Civil War (Score:4, Interesting)
This is a shot over the bow of the fab guys - there are alternates to you.
Given that GlobalFoundries licenses it’s fab tech (e.g. from Samsung), I don’t see how this is a shot at Intel’s fab guys. GF is doing great work, but they gave up on doing the R&D necessary to be on the bleeding edge a few years ago. By simply licensing it from others instead, they can save on risk and unknown costs, but they’re always a bit behind and aiming to make it up in volume. The strategy is working for them, but they aren’t an alternative to the work Intel’s guys are trying/failing to do on the bleeding edge.
Also, who spends $30 billion for a shot across the bow? This is an expansion into serving more customers, exactly in line with what they said they wanted to be doing.
Why Paywall (Score:5, Insightful)
Why do you bother to put up paywall source, then put up an alternative? Just put up the alternative and call it day.
Re: (Score:3)
Because paywalls don't affect everyone equally, while others prefer their sources from the mouth rather than the other end.
Re: (Score:2)
Then either put up two links or let the minority of readers who subscribe to paywalls click through to the original via reference.
Unless you come out in favour of a direct link to every scientific publication you're applying double standards.
Re: Why Paywall (Score:3)
The WSJ reported it, and only they know the source right now. So you're taking their word for it regardless which you read. Where did you think Reuters got the info from, the news fairy? Why shouldn't we have a link to the original then. Or if you're happy with any non original re-telling of a story, why don't you read the Slashdot summary and call it a day?
"... the Wall Street Journal reported on Thursday, citing people familiar with the matter."
Re: (Score:3)
So? There are many sites covering the story. If you want to click through to a paid access article then click through, but TFA should be accessible to all readers. Fuck post the Reuters story which has WSJ as the first link in it. But stop with this paywall bullshit.
Re: (Score:2)
Your uid is that low and you still can't figure out how to bypass a paywall? https://archive.is/Ys9WE [archive.is]
Re: (Score:3)
It's not a question of bypassing the paywall. It's we shouldn't have too. In case you haven't noticed, paywalled sources are not very popular around here.
I am waiting for Intel to screw this up (Score:4, Funny)
Like pretty much everything they've done lately.
Re: (Score:2)
Well we could have had in-chip ECC early if RAMBUS took off, instead of waiting for DDR5. [reddit.com]
Re: (Score:2)
RAMBUS was used as a monopoly tool by Intel, and thus was pricy for the consumer.
I can still remember a buddy from the MMO, RF Online, whom we had to wait for after each change of map while his PC was loading. (couldn't afford to extend the RAMBUS RAM)
Still a testament to RF Online efficiency:
RF Online, the mass PvP MMORPG that ran with 128 MB of RAM. Yes, Megabytes.
Re: (Score:2)
If by lately you mean the past 15 years or so, then yeah.
They gotta do SOMETHING (Score:5, Informative)
Which is why Intel is STILL using 10nm and their non-top-range products are even using 14... At the same time AMD is using 7nm and nVidia is using 5nm from TSMC. And TSMC is working with engineers to push 4nm into production this year while 3nm enters testing/sampling for next year.
Corollary of Moore's Law: The effective price of owning your own state-of-the-art semiconductor fab doubles every 18 months.
How bout no? (Score:2)
The sharks are circling (Score:2)
Time for the Government to put their money (Score:2)
where their mouth is. They talk a bit talk about tech monopolies, so they should just come right out and say they will block this move without further consideration.
Or perhaps they are hypocrites...
Nothing good for anyone will come from this purchase. Even shareholders will get screwed in 10 years when Intel writes this down due to their incompetence.
Re: (Score:2)
Oddly enough the government regulators that will ultimately approve this will "retire" shortly afterward and have board positions at Intel. Just wait and see.
Re: (Score:1)
They talk a bit [big?] talk about tech monopolies
It's campaign season, starts the day after the election
Intel's Director of Creative Innovation (Score:2)
In 2011 Intel named will.i.am Director of Creative Innovation. Really.
Maybe he can turn things around.
If not, they may have to sell that motorcycle they had Orange County Choppers build for them.
Seeing that the "computer" they installed on it was just an empty shell and never worked,
I doubt that they will get back anywhere near the $13 million they paid for it.
Welcome to Step 2. (Score:2)
My Post from 2008 [slashdot.org]
1) AMD Spins of fabs. ...
2) Intel/VIA/TMSC/IBM buys AMD Fabs.
3) Intel/VIA/TMSC/IBM Fabs charges huge price to manufacture AMD CPU's.
4) AMD CPU Prices skyrocket. Unable to find a cheap reliable FAB, AMD loses price competitive edge.
5) AMD Stock tanks.
6)
7) LOSS.
Welcome to Step 2
I suspect that Step 3 won't take nearly as long, and can argue that it's happening right now. To be fair, fabless has worked well for AMD so far with Ryzen, but with the chip shortage straining supply, and everyone usi
Re: (Score:3)
Re: (Score:2)
If anything is going to drive Step 3, it's going to be Fab capacity.
TSMC only has so much fab capacity. While they are planning to expand it's not going to happen in the short term and with big pockets (Apple, Nvidia, Intel, ETC) competing for FAB capacity, TSMC can basically name their price for FAB time. The only good thing here is that it's going to drive all chip prices higher, so at least AMD has less to worry about than in the original 2008 scenario. (where Intel exclusively made their own chips at th
Global Foundries should RUN away (Score:5, Informative)
I was an engineer in the Intel Digital Signage Group for about a month.
Management had the idea that they could dump the excess inventory of binned Core2 chips into the signage market
by telling potential customers that these were new devices designed specifically for billboard and kiosk applications. They were not.
The manager and his lieutenant were idiots. We were in Chandler AZ. There was an Intel group up in Hillsboro OR that was working on Code Signing Certificate tech.
Manager arranged a meeting with the group and had four engineers fly down on the Intel shuttle (Intel has its own employee airline, flying between CA, AZ, OR, and TX).
I told the manager repeatedly that Cert signing and Digital signage were not the same thing. He did not understand.
The meeting lasted for about 5 uncomfortable, embarrassing minutes and I was downsized out of the group the following week.
This is much of how Intel is outside of the x86 work. I had interviewed with the ill-fated, never-working Larrabee GPU group.
The project was killed before I got the nod and moved to Oregon. Thank goodness for that.
Re: (Score:2)
I told the manager repeatedly that Cert signing and Digital signage were not the same thing.
Random layperson here. What is the difference, as far as a CPU is concerned? I would have assumed that they just need to process some number of instructions per second within some TDP budget and environmental constraints.
Re: (Score:2)
The only thing cert signing and digital signage have in common is that they have nothing to do with x86 processor design. These three areas are about as far apart as you can get and still have them inside Intel.
Re: (Score:2)
"the ill-fated, never-working Larrabee GPU"
Larrabee = Phi, which was an actual product for a decade.
FTC will try to block (Score:2)
FTC will have a fit about Xeon and Epyc being owned by the same company. ARM servers just aren't there yet.
Intel knows this so maybe they know that the Biden folks would rather Intel own GF than the UAE. The FTC can be overridden, and sovereign chip capacity is important, but Intel will not be able to resist hurting Xeon competition if this goes through.
AMD can change but that's neither fast nor cheap.