CNBC: Thomas Caulfield, CEO of semiconductor manufacturer GlobalFoundries, negates reports by WSJ that Intel is in talks to buy the company in a $30B deal. "There's nothing to that story,"
Array ( [content] => [params] => Array (  => /forum/index.php?threads/intel-is-in-talks-to-buy-globalfoundries.14433/page-3 ) [addOns] => Array ( [DL6/MLTP] => 13 [Hampel/TimeZoneDebug] => 1000070 [SV/ChangePostDate] => 2010200 [SemiWiki/Newsletter] => 1000010 [SemiWiki/WPMenu] => 1000010 [SemiWiki/XPressExtend] => 1000010 [ThemeHouse/XLink] => 1000970 [ThemeHouse/XPress] => 1010570 [XF] => 2020770 [XFI] => 1050170 ) [wordpress] => /var/www/html )
But don't count US, European, and Japanese companies' commitment to Intel IFS and GF are guaranteed. Those potential customers' existing supply chains serve them very well, especially in terms of the cost. IFS/GF will be competing against many strong incumbents and new comers in the lower margin and matured process nodes.The 3 countries with biggest fabless customers are the US, Taiwan and China. Would US fabless use IFS-GF? Yes, although they could lose some customers that INTC competes with. Would Taiwan fabless use IFS-GF? No because they already have the ecosystem in Taiwan. Would China fabless use IFS-GS? No. Too much geopolitical risk. GF could potentially lose their current Chinese fabless customers. So IFS-GF would primarily be servicing US and European customers.
Any possible antitrust obstacles?The more I research this the more I like the idea. Intel recruited high level execs from GF and Samsung so I'm sure there are internal discussions happening at both companies. The people I have talked to inside and outside seem to think this is a great opportunity for Intel or Samsung to advance their IDM foundry business. It really will change the competitive landscape and $30B is not a lot of money considering fab costs and the expense of scaling the business side of a foundry.
Intel, TSMC and Samsung do not fear GF but if Intel buys GF Samsung will be at a huge disadvantage in the IDM foundry battle. The same goes for Intel if Samsung buys GF. Samsung would be unbeatable as the #2 foundry and #1 IDM foundry. No way does Intel or Samsung catch TSMC but it will certainly make the race more interesting.
From what I hear GF will do $6-7B in sales with most of their fabs fully utilized this year and they have plenty of room to grow in NY, Singapore, and Dresden. I quesstimate profits at $1B. I'm not an IPO specialist but I don't see a huge opportunity for a unicorn sized IPO here. We will have to wait and see what their filing says. They will definitely need to IPO before the semiconductor shortage narrative dies down so this year or next year at the latest.
Exciting times, absolutely.
The IPO valuation target from earlier this year was $20-30 B - for 2022 rollout - so this purchase price is in line with that. The existing fabs (Malta, Dresden & Singapore) would cost at least that much to duplicate, so the savings in time (to build) and development to get new customers would be accelerated 2-3 years - for either Samsung or Intel. The US-based and owned scenario is most likely to get more funding and support, so I'd bet that direction. The short term costs & profits will be ugly, but this is a long term play for both Intel and the US manufacturing base."#1 IDM foundry" ? Thought you didn't think "frenemies" worked in the foundry business. Why does it have to be an IDM foundry at all ?
I can absolutely see the logic for GF's owner's to sell at the top of the market. I'm less clear why that's a good deal for a buyer.
Certainly exciting times for investment bankers talking up both Intel buying GF and then perhaps spinning off the foundry business later.
Pat's comment on the Q2 call:
Hi, thanks. Pat, there were some headlines recently about you potentially looking at maybe building out your Foundry business by M&A, and I'm wondering, can you just comment broadly, do you think that M&A would significantly accelerate your Foundry efforts? I know right now you are basically offering the 22-nanometer process, and you probably have to offer more processes to pull those efforts forward. So I'm wondering if you can comment on the headlines that were out there. Thanks.
Hey, thanks for the question, Tim, and great to be with you all today. So first I'd say, obviously we can't comment specifically on the speculation that you've been hearing, but feel what I'm saying is, we're very happy with the build-out of the IFS business. As you say, it will include mature nodes, our 22 SFL, it will also include our leading edge nodes as well, our packaging offerings.t That overall, we're just seeing great momentum over a 100 customers in our pipeline. And we fully expect that this is going to be a great business for us.
At this point, we would not say that M&A is critical, but nor would we rule it out. Our view is that industry consolidation is very likely. The intense R&D, the need to move to modern and leading edge nodes, the massive capital investments required, we just simply view, that smaller players simply won't be able to keep up, and foundries without leading edge capabilities will be left behind. And we're continually seeking ways to accelerate our plans with IFS. If an acquisition can help, we will certainly not rule it out. Thank you.
So yes, Intel started the rumor, absolutely.
"and foundries without leading edge capabilities will be left behind."
Does Intel Foundry have a leading edge node process available to outside customers?
So for Pat Gelsinger to say "and foundries without leading edge capabilities will be left behind" is very strange. He paints a not so promising picture to his own Intel foundry strategy.Not really. IFS will be selling 22nm and 14nm wafers. Intel 10nm is backlogged with Intel products and Intel 7nm is 2-3 years out. TSMC 3nm will be in products starting in Q4 (Apple) so I do not see Intel ever having a "leading edge" foundry offering.