Major hedge fund advises Intel to outsource CPU manufacturing

[porina]

2 minutes ago, Master Disaster said:

What would probably end up happening is, whoever bought Intel's Fabs would end up running Intel CPUs until the end of a cycle before they could refit for other processes. Its not like Intel wouldn't still need silicon and from an economical PoV it would make little sense to shut down an operating fab to refit it while it still has a customer that is using it in its current state.

Maybe not in those words, but I thought that's what I said  Still, new fab owner would need to plan ahead for whatever is next. No easy job.

[straight_stewie]

21 hours ago, DuckDodgers said:

This is the second time this year Intel is pushed to reconsider its IC manufacturing capacity, but this one calls for a much more affirmative action, almost going fab-less.

I hope Intel won't vote to do this.

Buffet refers to what Intel is experiencing as "Mr. Market". Basically, the markets view of your company will change over the short and medium term against meaningless data.

Intel is still offering higher performance chips than the competition, and node size no longer matters. The only area where they are consistently behind at the moment is power consumption.

It is only a matter of time until the market realizes that AMD is tying meaningless numbers to their processor nodes and that Intel is still offering higher performance products for real workloads. Intel will come back.

Letting a venture capitalism firm do Soggy Cigar Butt investing on Intel (shedding assets for a quick flow of cash: "the last puff") would be a terrible mistake and would certainly bring about the permanent end of Intel being a serious competitor.

Especially since there are no other fabs that can handle Intels manufacturing needs (TSMC has already dedicated nearly the entirety of it's competitive manufacturing capabilities to Apple. Intel would have to pay them to build new fabs for them, so might as well just upgrade their existing fabs. It would be cheaper and would allow them to maintain control).

Intel is just as profitable this year as ever. Until that changes you'd be hard pressed to make a good argument that they should throw away the major portion of their business plan.

[whm1974]

I neither trust Hedge Funds nor Venture Capitalist Firms. Is it just me or so they tend to Ruin Companies outright?   

[porina]

22 minutes ago, whm1974 said:

I neither trust Hedge Funds nor Venture Capitalist Firms. Is it just me or so they tend to Ruin Companies outright?   

Only if they can make money for themselves in the process.

[Commodus]

6 minutes ago, whm1974 said:

I neither trust Hedge Funds nor Venture Capitalist Firms. Is it just me or so they tend to Ruin Companies outright?   

VC is essential to getting some companies off the ground, so I wouldn't trash them too much.

 

In this case, I do think Intel needs to swallow its pride and accept that it needs serious reforms, possibly with outside help. AMD and Apple are stomping Intel, and even 10nm desktop chips might not help for long.  There's a real chance the company that created x86 could lose all control of it... and if Apple has its way, could shrink x86's overall relevance.

[leadeater]

3 hours ago, straight_stewie said:

It is only a matter of time until the market realizes that AMD is tying meaningless numbers to their processor nodes and that Intel is still offering higher performance products for real workloads. Intel will come back.

This was more true for Zen 2 products, Zen 3 it's very much a no win situation for Intel now. Still fairly competitive but no real single better equivalent option now, you have to seek out situations like QuickSync to find cases where Intel can lead now.

 

Anyway yes Intel will be back, they weren't the market leader in fabs for decades for no reason. Intel got bitten by trying to be too good, a higher than their own typical gains. We'd all be having very different discussions if Intel had set lesser targets for 10nm and 7nm would be much closer.

[BuckGup]

19 hours ago, Master Disaster said:

Right but dumping the entirety of Intel's output on to TSMC and shutting down Intels fabs isn't going to help anyone.

It might help the investors lol

[inteli7.Ti]

Just now, BuckGup said:

It might help the investors lol

To be honest, I don't see Intel falling further. Rocket Lake is supposed to be on par with Zen 3, if not a little less in temps and multithreaded applications, using 10 nm rather than 7nm for Zen 3. Also, comparing nanometer sizes across multiple companies isn't very smart. IIRC, TSMC's 7nm is roughly the same desnity as Intel's 10 nm (link)

[charlie_root]

I've been an AMD fan for years because I wanted to see more than one major player in the market. Now that Intel is lagging behind, I really hope to see them bring their A game. I absolutely don't want centralization of processor fabrication, and especially not in the powder keg off the coast of China.

[porina]

4 minutes ago, inteli7.Ti said:

To be honest, I don't see Intel falling further. Rocket Lake is supposed to be on par with Zen 3, if not a little less in temps and multithreaded applications, using 10 nm rather than 7nm for Zen 3. Also, comparing nanometer sizes across multiple companies isn't very smart. IIRC, TSMC's 7nm is roughly the same desnity as Intel's 10 nm (link)

Rocket Lake is 14nm. Fair bet it is 18% IPC over Skylake since we know it is essentially same as Ice Lake architecture. That'll put it maybe between Zen 2 and Zen 3 overall. Intel 10nm wont arrive on desktop until Alder Lake.

[inteli7.Ti]

6 hours ago, porina said:

Rocket Lake is 14nm.

Sorry! I mixed it up with Ice Lake for their mobile lineup. So confusing 

[CommanderAlex]

We've already seen the outcome of a hedge fund and Sears, Roebuck & Co. Albeit Sears was already on the decline for a couple decades before but they just accelerated their decline. Lets not have that happen to Intel. 

[StDragon]

8 hours ago, charlie_root said:

I absolutely don't want centralization of processor fabrication, and especially not in the powder keg off the coast of China.

If that keg goes off, the price of used CPUs and hardware will go through the roof!!

 

New fabs need to be built in America and Europe ASAP!

[CommanderAlex]

1 minute ago, StDragon said:

If that keg goes off, the price of used CPUs and hardware will go through the roof!!

 

New fabs need to be built in America and Europe ASAP!

I'd love to see that happen, but IIRC (at least a 60 Minutes episode talking about contamination and recycling), China and some mine that has been closed in the US are the only places on Earth that have deposits of the rare earth metals that we need to produce electronics? Otherwise, we are relying off of exports of these metals in order to produce them elsewhere. 

[ezsteev]

19 hours ago, kaixax555 said:

I don't think Intel should follow the advice of hedge funds/activist investors/vulture capitalist firms.
Going fabless completely is a bad idea at this point, especially given that Intel is probably US semiconductor firm with the financial capability to compete with TSMC and Samsung.

Rather I think the investors should force a management shake up.
For a tech/semiconductor company, having a bean counter at the top is a bad idea.
Get engineers with good acumen to be at the top, have a solid roadmap for the future and push for fulfillment of the target objectives in the roadmap.

The sentiment above has no understanding of finance or business. This is underscored by trying to use a pejorative term "vulture capitalist" when Intel nothing close to a distressed entity.

 

Sell the fab, increase ROIC, lower your cost of capital, leverage up to buy back shares, use shares to buy startup with good ideas and employees.

 

This is how business works.

[charlie_root]

2 hours ago, StDragon said:

If that keg goes off, the price of used CPUs and hardware will go through the roof!!

 

New fabs need to be built in America and Europe ASAP!

IIRC TSMC was either looking into or already finalized plans to open a fab plant in Arizona. That's a start, but not enough.

[Lightwreather]

15 hours ago, charlie_root said:

I've been an AMD fan for years because I wanted to see more than one major player in the market. Now that Intel is lagging behind, I really hope to see them bring their A game.

This is actually how I've been feeling. I would never want either AMD or Intel go under the waters.

[Kisai]

10 hours ago, ezsteev said:

The sentiment above has no understanding of finance or business. This is underscored by trying to use a pejorative term "vulture capitalist" when Intel nothing close to a distressed entity.

 

Sell the fab, increase ROIC, lower your cost of capital, leverage up to buy back shares, use shares to buy startup with good ideas and employees.

 

This is how business works.

That is how bad businesses work. Good businesses do not take on venture capital, because they do not need it. Bad businesses that just want to sell themselves take on venture capital to cash out, and hope they aren't the ones holding the bag when the business goes under.

 

That is why the pejorative term "vulture capitalist" is used, because that is what happens every time venture capital is involved with start-ups and with companies that have sinking stocks. Some terrible investment people come in, carve up the company, take on more debt, and leave the carcass to rot.

 

There other kinds of investors (eg Berkshire Hathaway) which buy companies and then just run them without trying to turn them into something they aren't. Contrast that with companies run by clowns like Inspire brands (Arbys, Dunkin', Baskin Robbins,) Wendys (which was part of Arbys for 3 years),  Yum Brands (KFC/Taco Bell/Pizza Hut) and Resturant Brands International (Burger King and Tim Hortons) which keep shuffling their brands around in a brainless game of musical chairs of who's restaurants gets closed.

 

I keep seeing these companies get traded around like horses, and the end result for the consumer is higher prices and less restaurant choices as these Hedge funds run them into the ground.

 

 

[inteli7.Ti]

10 hours ago, Kisai said:

That is how bad businesses work. Good businesses do not take on venture capital, because they do not need it. Bad businesses that just want to sell themselves take on venture capital to cash out, and hope they aren't the ones holding the bag when the business goes under.

 

That is why the pejorative term "vulture capitalist" is used, because that is what happens every time venture capital is involved with start-ups and with companies that have sinking stocks. Some terrible investment people come in, carve up the company, take on more debt, and leave the carcass to rot.

 

There other kinds of investors (eg Berkshire Hathaway) which buy companies and then just run them without trying to turn them into something they aren't. Contrast that with companies run by clowns like Inspire brands (Arbys, Dunkin', Baskin Robbins,) Wendys (which was part of Arbys for 3 years),  Yum Brands (KFC/Taco Bell/Pizza Hut) and Resturant Brands International (Burger King and Tim Hortons) which keep shuffling their brands around in a brainless game of musical chairs of who's restaurants gets closed.

 

I keep seeing these companies get traded around like horses, and the end result for the consumer is higher prices and less restaurant choices as these Hedge funds run them into the ground.

 

 

I agree with what you're saying, but the semiconductor market is far different from more traditional businesses. The machines needed to make sub-7nm fabs cost hundreds of millions (i'm looking at you, EUV lithography) and the market is extremely undersaturated. 

[wanderingfool2]

1 hour ago, leadeater said:

And if you want to be truly different you'll most like have to contend with the OS and software support divide, an unquestionable death sentence to anyone and everyone.

Or the approach being be big enough, like Apple (which I know defeats the point still)

 

I agree with your statements though.

 

In regards to the topic as well, by having control of the chip manufacturing it means they can introduce new techniques that others might not have (or use).  e.g. the stacked nanosheet (while it might not even come to fruition the fact is they are still working on technology like that that sets it apart from other chip manufacturers)

[Gamer Schnitzel]

I don't care about Intel but they better not do this or there is going to be more demand for TSMC, etc. which already have barely any stock and then the GPU shortage will literally never end, literally. The thing a lot of people don't seem to understand is that companies do not just magically manufacture more because "they love money so so so very much". There is a very long process behind upping production. Companies want to own as few factories as possible while selling as much as possible because in the worst case scenario when demand decreases they need to sell those factories off or somehow turn them back into money which is completely impossible. The same happens when they have a new process or the product somehow changes and they need to buy new machinery. All of the old one has to go without any money being made of it. Unless TSMC or Samsung has God himself coming down to earth and promising them that they will 100% sell out for the next decade it is very unlikely they will start opening new factories to keep up with demand because they simply do not know whether this high demand will persist or not. This means higher prices for buyers and even higher prices for consumers. This is why Apple did not just buy one chunk of stock but 80% of the entire year's production. TSMC has to know beforehand how much demand there is. Intel and the others would have to sign a contract stating they will buy the next few year's of stock which again is risky for them since they can't predict the future so overall the consumer here is fucked. I hope Intel sticks to their own stuff. AMD has been the underdog for a very long time. I have zero problems with Intel being a joke for a couple of years, really no problems at all. People need to grow up and stop obsessing with "competition" in the CPU space. AMD literally has just gotten better performing products and people are already at it like rabid dogs begging Intel on their very knees to release better CPUs. Just chill and wait a year or two or more, who cares. Intel isn't going to die.

[leadeater]

18 minutes ago, wanderingfool2 said:

Or the approach being be big enough, like Apple (which I know defeats the point still)

To an extend sure, but they still do actually work with others even though they are one of the most independent of the bunch. But they only got like that from the benefit of being in it from the beginning and surviving, executing their vision. Very few were able to do that, not even IBM survived (they are a very different company now than before) in that sense.

 

But I doubt it'll get to the point where Apple is making their own displays, display protocols, networking and wireless, batteries etc. Even with Apple making their own CPUs/SoCs now they are/will rely on other technology partners and standards and have the additional benefit of having control over an established OS and software ecosystem.

 

Apple has a luxurious position that no other has or will ever get (probably). 

[Kisai]

3 hours ago, leadeater said:

 

Apple has a luxurious position that no other has or will ever get (probably). 

 

And that came from not treating the company like a traditional business. Had things continued after Scully was thrown out without Jobs coming back Apple would have likely stopped existing, and BeOS took over. The guy right before Steve Jobs came back went on to become a venture capitalist. Go figure.

 

It's one of those things that some people have no vision for the company in the position they've found themselves in and just kinda twiddle there thumbs as fires pop up, which I imagine is the situation with the Intel chip fab business. Something happened here, but it is not going to be fixed by dumping the fabs and outsourcing it. 

 

Like to present another analogy. What happens when businesses decide to sell their property and then just lease the very same property back? This other company is now responsible for the upkeep of the building, and is free to hire minimum wage janitorial and security staff, where otherwise when the company owned it, hired people who had the necessary training and legal reason to be on the property as they were employed directly.

 

Outsourcing never saves money, especially when you outsource the core business to Asia, because you then become beholden to logistics variables. It only ever makes sense to outsource non-core business, and even then only when there is no security threat in doing so. 

 

The amount of opportunity one has as a third party outsourcer to walk out the front door with the company secrets is lucrative, and just should not be done.

 

[porina]

1 hour ago, Kisai said:

It's one of those things that some people have no vision for the company in the position they've found themselves in and just kinda twiddle there thumbs as fires pop up, which I imagine is the situation with the Intel chip fab business. Something happened here, but it is not going to be fixed by dumping the fabs and outsourcing it. 

I don't think that is the case. They knew what they wanted to do, it just didn't work out.

 

I imagine the initial talks about 10nm went something like this:

 

Management: For 10nm, we want it to be the best, can you do all the good stuff?

Engineering: Well, there's many things we can do, each with their risks...

M: Do them all!

E: But there are risks that not everything will work... maybe we should take a safer route...

M: It'll be fine

 

Intel management probably wishes they could reload a save point and try a different story branch, but real life doesn't work that way. They just have to make the best of what they have now, and try to correct the course for the future, which they would have started years ago but it wont be entirely visible outside the company until it is the right time to announce.

[CarlBar]

37 minutes ago, porina said:

but it wont be entirely visible outside the company until it is the right time to announce.

 

This. I mean didn't AMD start Zen development fairly early into the bulldozer era, and look how long that took for us to get any hint of.