Thread regarding Intel Corp. layoffs

@19j+1jk40wztt Tariffs might take Intel down along with the rest of the sector, since it will take years to bring on new customers or even to bring more TSMC production in house.

$19 is not that special, imo, but breaking $17.50 could lead to a fast slide to the upper single digits. It was in the $7s after the dot com crash and then again in the 2008-2009 recession.

The market knows what bringing Caufield in as CEO is likely to mean, and he may clarify that pretty quickly for anyone who doesn't yet get it, so see how the stock reacts if he is the chosen one.

Not that Intel is going to the moon anytime soon but it could cause the stock to bounce around in the $20s some more. Be even more impressive if that happens as the rest of the sector profitability starts being eaten by tariffs. Intel is shaping up to be as contrary to the sector as ever.

@19k+1jk40wztt Chips Act won't stop a sale of some fabs.

The government might even be pushing for Intel to offload fabs and product groups, because x86 has such a large installed base that they don't want to see Intel fail, or fall further behind on the fab side.

The Board is always looking at alternatives but I'd agree that Caufield would be the perfect setup for breaking up Foundry, with Intel retaining 18A+ fabs. It would solve issues for both companies, and would have a greater chance of success if the CEO were from the acquiring company.

Intel would then be a GF customer for at least some of those older nodes, and that might help speed up transition to 18A.

Everyone associated with those fabs would become a GF employee and if so then they should expect a major headcount realignment over time, because no foundry has the excess headcount as does IFS.

IFS has way too many engineers and technicians, with tools being PM'd far too frequently. There are also too many tools and the yield expectations are too high for foundry. Much lower wafer cost when you stop trying to be perfect.

IIRC, Intel must maintain 51% control of the foundries due to chip act monies - so Intel (as a organization) can't go away. Doesn't mean it cant be morphed.

INTC loses this support ~$19, it's Timber Time

I expected Foundry to be broken up in 2025 and if Caufield gets the CEO job then that would almost have to be in order to facilitate selling all fabs which are not EUV capable (or maybe all fabs not ramping 18A) to GF.

For companies in transition, CEOs are hired because they either have been through a similar transition, have developed a network which is needed for the transition, or (in this case) previously led the company which will be doing M&A.

Intel Foundry and GF probably would have a hard time getting a merge approved and Intel probably wants to keep everything from 18A forward. They have not even set up the older nodes to take external customers.

So they sell all older nodes to GF, then buy the wafers from GF.
This enables GF to have a profitable operation while they start to bring in external customers.
Intel is able to shed about 25% of the workforce and clean up the balance sheet, with more than enough to fund Foundry till it picks up a sustainable customer base.

The stock probably becomes stronger than the sector or at least is on par with the sector (x86 is still losing to ARM). As other business units are spun off or sold that helps too.

Andy Grove referred to this as a revolving door, where you walk out of the building then walk back in, unencumbered by the past.

This is how tech companies pull out of the ditch. Where it goes from there is TBD.

@e2+1jk40wztt That does sound fishy.

I think the Board was naive or deluded about how quickly they could bring on customers, and also about the willingness of any customer to use Foundry when it was part of the same company as the Product groups, a conflict of interest waiting to happen.

Until they spin off Foundry, I don't think customers are going to accept doing business with Foundry which is still in any way part of Intel.

Seems like remnants of the old arrogant Intel, which need to just go away. It wasn't that they were ever that good, they simply had a monopoly for a decade or so.

Even before foundry expansion the existing fabs were not fully utilized.
How did the leadership and board decide the future Mega foundries are going to be fully occupied?
The brings the scenario like this.
The fishing company with one boat was not profitable to catch enough fish to pay off the boat and labor.
Now the fishing company decided to buy more boats with loan from others and hire more fishermen expecting they are going to catch more fish to be profitable.
This is not scalable at all.
Did the leadership think even that they have enough fish in the fishing area?
Now the company left with lots of boats and fishermen without any earnings.

@a6 remember the Yugo?

@bg+1jk40wztt The Board seems to be treating Products as mere cash flow to build up Foundry. Product groups can downsize as needed to keep that going a long time.

The real (remaining) value is in Foundry, and OP is probably right that even the more recently scaled down capital plan is still too fast of a ramp for the rate at which external customers are onboarding. It is so far one at a time, and plenty of existing and near term capacity to handle that.

Any obsolete fab shells from legacy nodes could just be shut down and left idle but the company may find a way to sell or lease them as equipped fabs, even though they are all on campuses which also have newer fabs. That would be 'smart' capital.

The WSPW capacity of each fab is sufficient to be profitable, and that has increased over time as needed. So the only reason to add more fab capacity is to meet proven demand, and Dave actually used those words so it is clear enough that ELT gets it.

@a6+1 You live in Intel La La land. Good luck!

that and terribly designed products ..

Don't know this for sure but the oldest fabs may not be able to handle the requirements for the EUV school bus. It is hugh..

The tariffs (as discussed so far) are for finished goods, so that might drive some packaging to Intel and that has been a good business already.

They could change the tariff scope for semiconductors but so far they have not.

What you won't see is a continuation of corporate welfare, because it does not change the current customer incentive to remain at TSMC.

The company has already scaled back the fab expansion, which is all for newer nodes. The technology has gotten to the point where new nodes have larger scale requirements to be profitable but I think even with the current fab expansion there is more capacity that any amount of the current external customer onboarding will need.

The issue is with the older node fabs, because those nodes are not profitable for the kind of products Intel wants to produce, and the cost structure is too high for any external customers to use those nodes. This is due to the patterning requirements needed for client and server products. Global Foundries or some smaller foundry might have a use for them in automotive or industrial.

It is those older fabs that appear to be holding up any takeover, and cause issues for plans to spin off Foundry. The company has been taking impairments related to this essentially stranded capacity and the current plan is to just keep eating losses till 18A ramps enough to shut down those older nodes towards the end of the decade.

The problem is that onboarding external customers is a slow process and those customers don't even start till 18A is up to production. It is the realization of this timing issue that caused the Board to shift strategy.

They now are breaking up the company and selling off everything needed to get through this multiyear period between the foundry ramp investment and when they can get meaningful external customer revenue.

Selling some of those older node fabs would go a long way towards closing that gap, and would make Foundry far more viable to an external investor or takeover.

I'd expect if somehow the company does get taken over, selling off the older fab assets would be part of that planned acquisition.

It will be a nice little laptop cpu supplier. Nothing more, nothing less.

Foundry will have to get smaller

No.

In order to operate a leading edge node, you need to be gigantic. You need monster units to recover the large up front costs. That’s why fabs consolidate to only a handful of winners. At least for the leading edge nodes. Economies of scale. This is also why Intel can’t develop leading edge on x86 volumes alone.

Trailing edge can afford to be smaller.

Tariffs will shift manufacturing to US. Trump is not going to let Intel, a national hero for semiconductor, die. Global companies won't let TSMC have a monopoly either. Each company giving 5% of their orders to Intel will help through the transition.

Printing money or typing numbers on a screen to generate money is easy. Building real technology is not and this is ultimately the real credibility behind the power of any currency. The credibility of the government to generate real value through taxes based on technology, raw materials, labor, capital assets, etc.

Change CHIPS act to give to domestic companies AKA Intel and make it Americas preferred semiconductor manufacturer.