News Intel 2Q24 Financial Results

[ErichZannMusician]

Every semiconductor company has had 15-20% stock price wipeout since recent highs, Blackwell is delayed by at least 3 months, Nasdaq is heading for a conservative 800-1000 pt correction in the short term - maybe in August itself, rate cut has been announced too late causing panics of slowdown, $70bn fund managers are cautioning against the AI hype.

And amidst this Intel's CCG has carried them all along with 30% operating margins.

Intel will be more than fine in the long term.

 

[jpiniero]

CCG ain't gonna pay the fab bills.

 

[ErichZannMusician]

CCG ain't gonna pay the fab bills.

They don't need to.

Intel is trying to make Foundry stand on its own. It is the only way the US gets its own TSMC, on US soil.

And Intel is shopping for talent from TSMC.

 

[DAPUNISHER]

Here is a week wrap up on tech stocks earnings. Covers Intel, Qualcomm and AMD.

https://twitter.com/x/status/1819373742579208625

Something that jumped into my head while watching that is I'd much rather be selling Microsoft Instinct than semi-custom chips for consoles. Also, Intel is boned; that line about his takeaway was come see us in 2030? I'm dead. 🤣

 

[jpiniero]

Intel is trying to make Foundry stand on its own.

And failing miserably.

 

[gdansk]

Every semiconductor company has had 15-20% stock price wipeout since recent highs

Yes, it's an industry-wide problem.





These all look the same to me.

 

[DAPUNISHER]

They don't need to.

Intel is trying to make Foundry stand on its own. It is the only way the US gets its own TSMC, on US soil.

And Intel is shopping for talent from TSMC.

You have not seen what Intel just announced I take it? If you watch the vid Joe provided, the amount of money the feds gave Intel was not a resounding endorsement. Recruiting quality talent during layoffs and axing part of the benefits package is a bold strategy Cotton...

 

[ErichZannMusician]

Yes, it's an industry-wide problem.
View attachment 104387

View attachment 104388

View attachment 104389
These all look the same to me.

"recent highs"

 

[ErichZannMusician]

You have not seen what Intel just announced I take it? If you watch the vid Joe provided, the amount of money the feds gave Intel was not a resounding endorsement. Recruiting quality talent during layoffs and axing part of the benefits package is a bold strategy Cotton...

Intel is hiring from TSMC in Arizona.

Wasn't one one of the complaints that Americans had was the work-culture at TSMC?

 

[gdansk]

"recent highs"

I'm genuinely curious, can you find a time frame where Intel doesn't have the biggest drop?
I was trying to find one but couldn't.

 

[DAPUNISHER]

Intel is hiring from TSMC in Arizona.

Wasn't one one of the complaints that Americans had was the work-culture at TSMC?

Intel is laying off almost 20K employees but they are hiring? Explain please.

 

[gdansk]

Intel is laying off almost 20K employees but they are hiring? Explain please.

Well this does happen all the time.

 

[DAPUNISHER]

Well this does happen all the time.

Not good enough. It needs factual support. Intel is in a death spiral at the moment; recruiting isn't conducive to cost savings via layoffs. Talent is not going to work for less than the departing employees is my hot take, happy to be proven wrong.

 

[Failnaught]

I lost track: Lunar Lake is made by TSMC. What is the next CPU fabbed on an Intel process and when is it coming? I suppose this will give a (optimistic) indication of how confident Intel is in their own manufacturing.

Whatever happened to their GPU line? Are they close to competitive, and what is the barrier to success? Is it the software support (CUDA)? Even AMD is having trouble penetrating into AI market.

 

[alcoholbob]

I lost track: Lunar Lake is made by TSMC. What is the next CPU fabbed on an Intel process and when is it coming? I suppose this will give a (optimistic) indication of how confident Intel is in their own manufacturing.

Whatever happened to their GPU line? Are they close to competitive, and what is the barrier to success? Is it the software support (CUDA)? Even AMD is having trouble penetrating into AI market.

Panther Lake, which is supposed to replace Lunar Lake mobile CPUs in 2025 is on Intel 18A. I don't think we know what the next generation desktop line is, it was supposed to be Beast Lake originally but its been reportedly cancelled so what comes next is...(?)

 

[Nothingness]

Not good enough. It needs factual support. Intel is in a death spiral at the moment; recruiting isn't conducive to cost savings via layoffs. Talent is not going to work for less than the departing employees is my hot take, happy to be proven wrong.

It depends how deep the restructuring is: if they completely get rid of some divisions/business units, they can lay off all the people including highly paid talented and experienced (dare I say "boomers" 😜) people. This could allow them to hire people for the parts they want to concentrate on, while still saving money.

But if they reduce salaries, they will have a hard time attracting talents. Or even worse they could get people that were rejected from all other companies because they were not up to the task. Getting talented people has been very hard for years and years.

I guess we'll have to wait for Intel to give more information about their plans, but IMHO if they don't recruit fresh blood with new ideas, their won't get out of their death spiral. And even more important, they need to change their internal culture, and there, I agree, it's a place where old timers might become a problem including among engineering.

 

[AnandThenMan]

I don't know about that. I bet the average Joe still walks into Best Buy and has no problems with buying a computer with an Intel sticker on it. If their reputation is damaged by the 13th/14th gen issues enough where even average Joes ask sales reps in Best Buy to show them non-Intel computers only, that's when it will hurt.

In summary Intel's business model relies on the ignorance of their customers.

 

[ErichZannMusician]

I lost track: Lunar Lake is made by TSMC. What is the next CPU fabbed on an Intel process and when is it coming? I suppose this will give a (optimistic) indication of how confident Intel is in their own manufacturing.

Whatever happened to their GPU line? Are they close to competitive, and what is the barrier to success? Is it the software support (CUDA)? Even AMD is having trouble penetrating into AI market.

Intel's products are finally no longer tied to their own nodes.

Look up Lion Cove presentation to see how Intel is on its journey to become process node agnostic.

With that being said, Intel is placing the future of the foundry business on 18A - which is their own node.

There are two future products fabbed on 18A - Panther Lake for consumer and Clearwater Forest for DC.

Panther Lake is supposedly up and running Windows and Clearwater Forest is on track.

Panther Lake is based on the Cougar Lake P-core design, a minor bump from Lion Cove.

Clearwater Forest is based on Darkmont - an evolution of Skymont.

 

[Failnaught]

Designing a real, serious product on TSMC might even be a good idea: you get flexibility if the fab really doesn’t work out, and you gain valuable experience working with a foundry as a customer. If you rotate some of those design engineers to the foundry side, they will better understand what their customers need.

I guess a one product gap in their lineup is ok, but now the pressure is on for 18A! I’m sure they’ve got their best people working on it!


Sent from my iPhone using Tapatalk

 

[Tup3x]

Not good enough. It needs factual support. Intel is in a death spiral at the moment; recruiting isn't conducive to cost savings via layoffs. Talent is not going to work for less than the departing employees is my hot take, happy to be proven wrong.

If Intel is doomed... that's not going to be a good thing for x86.

 

[Timorous]

If Intel is doomed... that's not going to be a good thing for x86.

AMD should buy intel then let others use x86...

 

[jpiniero]

Not good enough. It needs factual support. Intel is in a death spiral at the moment; recruiting isn't conducive to cost savings via layoffs. Talent is not going to work for less than the departing employees is my hot take, happy to be proven wrong.

It's extremely common for layoffs to really be about getting CHEAPER employees and not less.

However in this case I have seen people claim that there is a hiring freeze. And thus any job postings are simply fake.

 

[DAPUNISHER]

It's extremely common for layoffs to really be about getting CHEAPER employees and not less.

However in this case I have seen people claim that there is a hiring freeze. And thus any job postings are simply fake.

Those jobs are not coming back anytime soon.

 

[dmens]

But if they reduce salaries, they will have a hard time attracting talents. Or even worse they could get people that were rejected from all other companies because they were not up to the task. Getting talented people has been very hard for years and years.

FWIW, at least within hardware engineering, Intel is already the lowest paying big company on average. The PC resurgence during COVID saw a brief period where Intel started making marginally competitive offers but that is totally over.

The real potential catastrophe is Intel laying off the already relatively cheap labor who have been there for a long time and are paid way under market due to low starting pay and no market adjustment. From a financial perspective, those people are the best deal Intel will ever have, and are absolutely irreplaceable for the same unit cost. It remains to be seen if Intel recognizes this. It most certainly did not in the last round of layoffs in 2016.

The above may be a moot point if the plan is actually to kill off large numbers of products, or if there is an actual plan to massively increase the efficiency of their engineering orgs. I am skeptical of both.

 

[igor_kavinski]

Look at this lovely MTL bug:



That BIOS fix will probably be in the form of a prevention from running at max turbo frequency. And since this BIOS fix is not implemented yet, a turbo boosted MTL CPU could fail to thermally throttle and possibly DIE from high heat exposure

And they spent who knows how much exactly on hot lots and incurred massive losses trying to supply this broken silicon to OEMs!