News Investors pushing Intel to split off manufacturing

NTMBK

Lifer
Nov 14, 2011
10,208
4,940
136

(Reuters) - Activist hedge fund Third Point LLC is pushing Intel Corp to explore strategic alternatives, including whether it should keep chip design and production under one roof, according to a letter it sent to the company’s chairman on Tuesday that was reviewed by Reuters.

Were it to gain traction, Third Point’s push for changes could lead to a major shakeup at Intel, which has been slow to respond to investor calls to outsource more of its manufacturing capacity. It could also lead to the unwinding of some of its acquisitions, such as the $16.7 billion purchase of programmable chip maker Altera in 2015.

I wonder if Intel will take heed? They certainly need to change something.
 
  • Like
Reactions: krumme

Jimzz

Diamond Member
Oct 23, 2012
4,399
190
106
In the short run it would help the stock. But look at AMD. They can't get enough chips as is and they are still small fries compared to intel.

Only 2 things would happen in the long run and both are not good for intel.

1: The spun off Fab falls on its face and can't get any better nodes up and running. Intel is now competing with AMD, Apple, etc for fab space and can't even begin to get enough chips and hurts them. Then on top of that the old Intel fab now makes chips for others so intel has to get in line there and pay higher cost per wafer.

2: The spun off fab gets its act together and catchs up with other Fabs. But now they sell to the higgest bidder so they have to fight with AMD, Apple, etc... for wafers and pay a higher price than before.



So the only good thing that would come of this is the investors would get a large jump in value in the short run. But long-term would hurt and maybe even kill intel as we know it.

Instead how about getting better management. Intel keeps moving the deck chairs but seems the same type of people and ideas are still there. Look what happened when AMD cleaned out a lot of legacy management.
Intel still has a lot of weight and money. There is no reason that with proper managment they can't retain their ability to compete if not dominate.
 

shady28

Platinum Member
Apr 11, 2004
2,520
397
126
Hedge funds.... should be outlawed.





 

shady28

Platinum Member
Apr 11, 2004
2,520
397
126
Remember when Intel's "Ace in the hole" was that they had the fabs?
Only the bean counters concerned about the next earnings report would consider this. They'll jump ship once they cash in.

Exactly. Sell the fabs, do a big dividend and stock buyback, then the hedge funds will sell. When they're done, nothing but the wreckage and shell of the former company is left. This is the pattern of "activist investor" hedge funds.

Hedge funds exist in the first place through loopholes allowing them to dodge regulatory oversight, while getting tax benefits as if they were real capital investment companies. In the real world they act as corporate raiders. That needs to be fixed.
 

dmens

Platinum Member
Mar 18, 2005
2,271
917
136
do a big dividend and stock buyback

LOL! So it is bad when the evil hedge funders supposedly plan on doing this after seizing control, but it is totally fine when the Intel CEO and board do exactly that for the last four years to prop up the stock?

The open letter did not even tell Intel to sell the fabs, so why are you so incredibly butthurt? Can you actually come up with any real criticism on the points in this letter other than "muh evil hedge funds"?
 
Last edited:

dacostafilipe

Senior member
Oct 10, 2013
771
244
116
Well, I don't think that Intel manufacturing is structured in a way to "earn" money, so this would fail without massive investment from external sources. GloFo all over again ...

If Intel really lost the ability to produce state of the art manufacturing, just sell your Fabs to TSMC with some kinda of exclusivity contract, focus on your architectures and give us top CPUs again ... please!
 

NTMBK

Lifer
Nov 14, 2011
10,208
4,940
136
I don't know how well Intel fabs would do as a merchant foundry. Intel already tried to have a foundry business, and it basically killed LG's SoC plans- look up the Nuclun 2. It was meant to be on Intel 14nm, but it never came to market. Their other high profile foundry customer, Altera, ended up getting bought by Intel.
 

Hulk

Diamond Member
Oct 9, 1999
4,191
1,975
136
Well, I don't think that Intel manufacturing is structured in a way to "earn" money, so this would fail without massive investment from external sources. GloFo all over again ...

If Intel really lost the ability to produce state of the art manufacturing, just sell your Fabs to TSMC with some kinda of exclusivity contract, focus on your architectures and give us top CPUs again ... please!

Or instead of throwing your hands up in the air and quitting do it the old fashioned way.

Roll up your sleeves and get to work. The board should get people in charge at the top that understand the problems and know (or have the smarts) how to fix them. Bring in the appropriate people at the fabs, offer bonuses for milestones, you know, lots of late nights, hard work, and a desire to be the best.
 

Mopetar

Diamond Member
Jan 31, 2011
7,797
5,899
136
Considering it's the same board that let the company get to this point, is there really any indication that they are capable of fixing the problem or that they aren't also a part of it?

It sounds like this all is the result of a lot of bad policies and practices that were able to build up over time because the company didn't have any real competition. Fixing that isn't just as easy as getting the right people hired if they're hamstrung by bad policy.
 

Doug S

Platinum Member
Feb 8, 2020
2,201
3,405
136
The time for Intel to spin off their fabs would have been 10 years ago when they were ~3 years ahead of everyone else. That would have been a huge IPO, and there would have had a lot of interested customers knocking on their door, and they could charge a premium. Now that they are behind, who will be their customer base other than Intel's own designs (the ones that aren't moving to TSMC to access a leading edge process, that is) and US military/government stuff that must be made in the US for security reasons or due to "buy in the USA" laws.

They would be forced to compete on price with one foundry (TSMC) who will have already greatly depreciated competing processes thanks to Apple & Qualcomm, and another foundry (Samsung) who has lots of extra capacity and likely hungry for more market share. Good luck with that!
 

dmens

Platinum Member
Mar 18, 2005
2,271
917
136
Where did you get that sentiment from the post?

Whining about hedge funds supposedly advocating for stock buybacks even though this particular case did not, and advocating for the outlawing of hedge funds for this, without mentioning it is common practice for dead-ender company boards to buy back stock. Especially when the company in question publicly announced an accelerated stock buyback just a few months ago.
 
Last edited:

scannall

Golden Member
Jan 1, 2012
1,944
1,638
136
Or instead of throwing your hands up in the air and quitting do it the old fashioned way.

Roll up your sleeves and get to work. The board should get people in charge at the top that understand the problems and know (or have the smarts) how to fix them. Bring in the appropriate people at the fabs, offer bonuses for milestones, you know, lots of late nights, hard work, and a desire to be the best.
IBM gave Global Foundries a *working* 7nm process. They had some of the best process engineers on the planet. But, the company decided to not spend the money on the equipment. Intel could have poached great talent at that point, but didn't try. So others did. And that talent in now likely unavailable.
 

NTMBK

Lifer
Nov 14, 2011
10,208
4,940
136
IBM gave Global Foundries a *working* 7nm process. They had some of the best process engineers on the planet. But, the company decided to not spend the money on the equipment. Intel could have poached great talent at that point, but didn't try. So others did. And that talent in now likely unavailable.

A process suitable for IBM's needs is not necessarily suitable for the rest of the semi market. The prices on IBM's big iron mainframes are insane, so I imagine that they can eat huge per-wafer costs and low yields.
 

scannall

Golden Member
Jan 1, 2012
1,944
1,638
136
A process suitable for IBM's needs is not necessarily suitable for the rest of the semi market. The prices on IBM's big iron mainframes are insane, so I imagine that they can eat huge per-wafer costs and low yields.
I'm going to respectfully disagree here. The IBM Power9 and well all the other Power series parts are huge.The yields on much smaller parts would be more than good enough. IBM more than understands high speed. They were the ones who released the highest clocking CPU's at stock than anyone. (5.5 Ghz zEC12)

As far as GloFo not buying 7nm equipment. It may have just been a too expensive thing. Or it may have been the line was so long they may have thought they wouldn't get it soon enough to make money at it. But, looking at TSMC, they really should have.
 

Hulk

Diamond Member
Oct 9, 1999
4,191
1,975
136
One thing we can't overlook here is the fact that the resources that are put into the research, development, and finally production of any product is directly related to how many of that product can expect to be sold (size of the market), predicted longevity of the market, and price of the product.

A great example is the smart phone. New to the world in 2007. They lit the world on fire. 15 year ago no one had one, now everyone has one. Trillions of dollars have driven that market and the result is we have high powered computers with large storage and high quality, high resolution screens in our hands. Look at the ground ARM-based processors have made on x86 over the past 13 years. It's astounding and driven by an enormous amount of sales and competition.

There is no doubt there has been a decline in the desktop computer and a move to portable devices. For many people all they need is e-mail and internet. And the smart phone will do that.

We've seen the shift Intel has made over the last 5 years or so to prioritizing mobile processors over desktop. And despite the fact that AMD has a very competitive architecture and more than competitive process they still haven't really "landed" with force in that market. For all of Intel's woes it did manage to get Tiger Lake out, which of course has a very competitive core and somewhat competitive 10SF process.

My point is that Intel may be losing faith in the importance and profitability of the desktop CPU and is unwilling to spend the resources they might have 15 years ago.
 
Last edited:
  • Like
Reactions: Failnaught

TheELF

Diamond Member
Dec 22, 2012
3,967
720
126
My point is that Intel may be losing faith in the importance and profitability of the desktop CPU and is unwilling to spend the resources they might have 15 years ago.
PC centric is the biggest segment for intel and while that is not just desktop it's a big part of it.
14nm still works and still makes them huge amounts of money so why fix something that isn't broken? The only drive companies have to make better products is if they have to make more sales.


Also intel makes more money than tsmc amd and nvidia put together and if their upcoming GPUs are halfway acceptable they are going to take away a lot of nvidia's and amd's marketshare in GPUs.
It makes a lot of sense that intel does a lot of stock buy back before the release of them.

 

Gideon

Golden Member
Nov 27, 2007
1,608
3,571
136
Also intel makes more money than tsmc amd and nvidia put together and if their upcoming GPUs are halfway acceptable they are going to take away a lot of nvidia's and amd's marketshare in GPUs.
It makes a lot of sense that intel does a lot of stock buy back before the release of them.

Le'ts look at a more recent report from the same site you listed. Sometimes it's more important to look at the longer trends rather than what's the deal now:
Fabless/system IC company sales are forecast to more than double from 2010 to 2020 ($63.5 billion to $130.0 billion) whereas the total IDM IC sales are expected to be up only 30% over this same timeperiod, from $204.3 billion in 2010 to $265.7 billion in 2020.

1609496973600.png

And before anyone starts drawing conclusions about Intel and the decline in 2017-18:
In 2002, fabless/system IC company sales accounted for only 13% of the total IC market. With the memory market soaring in 2017 and 2018, a market in which the fabless companies have very little share, the fabless share of the total IC market shrank in both of those years. However, with the memory market registering significant weakness last year, this situation reversed itself, with the fabless share of the total IC market increasing 3.9 percentage points in 2019 to 29.7%.

Some additional points:
  1. Intel's client GPUs will be built in other fabs, adding to their bottom line not Intel's.
  2. At least some of the revenue you listed came from Intel's memory business they just sold off to SK Hynix.
 

Gideon

Golden Member
Nov 27, 2007
1,608
3,571
136
Let's also look at things the other way.

Just 2-3 years back TSMCs foothold in Intel's core markets was in all essence 0. Now we have:
  • AMD having 20.2% of the entire Q3 x86 CPU market share (including 20% in laptops, yet only 6.6% in servers) in Q3, most of it being TSMC (and remaining scraps GloFo).
  • Apple ditching Intel ( another 7-8% of laptop market-share but gaining ground) - all TSMC
  • Amazon Graviton release (I can't dig up the article but AFAIK 20% of instances are already Graviton and another 20% Rome).
  • Altra Ampere release - all TSMC

Considering Google and even Microsoft are also readying their own ARM servers I'd say Intel's share in the Total Addressable Market is certainly going down in the mid-long term. Hyperscalers count to about half of the entire server market.


As for 20Q4 and 21Q1:
  • Zen 3 release certainly means improved market share in desktop Q4 i mean, just look at mindfactory numbers
  • With Cezanne in more laptops, including high-end (where it's never been before) you can expect Q1 to improve even more as it looks that Rocket Lake and Tiger Lake-H will only have any impact in Q2 and beyond.

I'm sure Intel will reverse some trends in desktops in 2021. They might stabalize in laptops but I doubt they'll be able to completely reverese it (as Apple is pivoting away on top of AMD is being competitive, Windows on ARM might also rear it's ugly head). And servers is a really long game that looks to have very little positive for intel in the long run. Hyperscalers have vested interest to reduce their share massively, even when Sapphire Rapid ends up crazy-good.
 
Last edited:

jpiniero

Lifer
Oct 1, 2010
14,509
5,159
136

I dunno, to me it looks like AMD is selling a ton of Picasso. It also looks like to me that TSMC is unable or unwilling to expand to the point to be able to completely own the market. A large portion of the PC market is going to need to be fabbed on something other than TSMC's latest just simply due to capacity.