News Intel 2Q24 Financial Results

[DavidC1]

I have no great love of Intel and their business practices but having two strong competitors bodes well for consumers IMO.

Looking at the state of Intel now, and the fact that is despite them still having 80% marketshare in client which is the only part propping them up, how are we going to have two "strong" competitors?

Remember Ivy/Haswell/Broadwell/Skylake? They have process leadership and barely got out 10% gains every two years, while the clock speeds didn't increase much and all were 4 cores?

Did they acquire some magic to get nearly 20% in the past few years despite immense struggles with process or did they purposely make decisions not to invest in their product lines?

Intel is every bit as monopolistic and greedy as Nvidia WITHOUT the execution part. At least we'll have better CPUs with them.

 

[DAPUNISHER]

Intel is every bit as monopolistic and greedy as Nvidia WITHOUT the execution part. At least we'll have better CPUs with them.

100% this.

As to the question of why Nvidia would want to introduce a hardware solution when they offer cloud gaming: Cloud gaming is no good for competitive titles. And how can you be like Apple if you don't have your own hardware ecosystem?

 

[DAPUNISHER]

Imagine the resources Nvidia has at their disposal to throw at CPU development if they buy Intel. We might get some exciting products initially but, once the juggernaut has its momentum, I think they would crush AMD or at least relegate them to a tiny fraction of the marketplace for both GPU and CPU. Then we get either super expensive products like we have currently on the GPU side or we get pathetic performance uplifts like we got for years under Intel when they dominated the market.

Analyst say Intel brings poor value and Nvidia would never buy them, even if the acquisition were permitted.

I have no great love of Intel and their business practices but having two strong competitors bodes well for consumers IMO.

This is another oft repeated trope that is quickly becoming outdated. Snapdragon is only the first salvo. Desktop and mobile are no longer a 2 horse race.

Qualcomm wants some of the pie - https://www.windowscentral.com/hard...e-desktops-powered-by-snapdragon-x-processors

Intel is fighting hard battles on every front. They are quickly running out of the talent and money to to do it successfully. No longer are they positioned to win a war of attrition.

 

[Elfear]

Looking at the state of Intel now, and the fact that is despite them still having 80% marketshare in client which is the only part propping them up, how are we going to have two "strong" competitors?

Remember Ivy/Haswell/Broadwell/Skylake? They have process leadership and barely got out 10% gains every two years, while the clock speeds didn't increase much and all were 4 cores?

Did they acquire some magic to get nearly 20% in the past few years despite immense struggles with process or did they purposely make decisions not to invest in their product lines?

Intel is every bit as monopolistic and greedy as Nvidia WITHOUT the execution part. At least we'll have better CPUs with them.

Analyst say Intel brings poor value and Nvidia would never buy them, even if the acquisition were permitted.


This is another oft repeated trope that is quickly becoming outdated. Snapdragon is only the first salvo. Desktop and mobile are no longer a 2 horse race.

Qualcomm wants some of the pie - https://www.windowscentral.com/hard...e-desktops-powered-by-snapdragon-x-processors

Intel is fighting hard battles on every front. They are quickly running out of the talent and money to to do it successfully. No longer are they positioned to win a war of attrition.

I should clarify that I'm coming from the perspective of a desktop gamer where frankly Snapdragon and Apple aren't a part of the picture. For gaming, it's essentially a two pony race on the CPU front and a stallion, pony, and 2-legged dog on the GPU front. I get what you and @DavidC1 are saying in that if Intel can't turn things around it would be better for Nvidia to buy them up and execute in a much more efficient and effective manner. I'd rather see that scenario too (although it sounds like Nvidia acquiring Intel is a non-starter).

My main point was that AMD has a better chance of remaining a strong competitor against a resurgent Intel on the x86 side than it would against Nvidia entering that market. Just my .02.

 

[adroc_thurston]

than it would against Nvidia entering that market

NV is utterly incompetent at all things CPU and SoC.
QC kicked them around like toddlers in phone SoC space in early 10s. Wasn't even a fair game!

 

[Hitman928]

Looks like Intel isn’t counting on any major customers for 18a, at least not at “launch”.

Intel will begin to generate a "meaningful" amount of revenue from its contract chip manufacturing business in 2027, the company's CFO said at an investor conference on Wednesday.
Intel is currently in talks with 12 potential customers that finance chief David Zinsner said will generate some revenue in 2026 and additional cash in 2027.

https://www.reuters.com/technology/intel-manufacturing-business-will-see-meaningful-revenue-2027-cfo-says-2024-09-04/

 

[DAPUNISHER]

NV is utterly incompetent at all things CPU and SoC.
QC kicked them around like toddlers in phone SoC space in early 10s. Wasn't even a fair game!

I know this one, it was an ad in my comic books. Just like the protagonist, Nvidia has been getting swole, and the rematch will be very different.

 

[Joe NYC]

I don't think that Nvidia would be allowed to buy Intel.

NVidia is already being investigated by DOJ for anticompetitive practices.

Allowing 2 monopolists to merge? Agreed, it is not going to happen.

 

[Joe NYC]

Intel announces cancellation of 20A process node for Arrow Lake, goes with 'external nodes,' likely TSMC​

Not that long ago, some Intel faithful were in denial that one (of 2) CPU tiles, the high end one would be on TSMC. They believed all would be Intel 20A.

And let's not forget, the 8+32 Arrow Lake on 20A they just "knew" was coming...

Now, Arrow Lake is all TSMC.

But, on the bright side, Intel foundry got a "packaging win", from Intel itself, to package all the TSMC die.

Intel announces cancellation of 20A process node for Arrow Lake, goes with external nodes instead, likely TSMC [Updated]

Out with the Intel node, in with TSMC.

www.tomshardware.com

 

[techjunkie123]

Intel announces cancellation of 20A process node for Arrow Lake, goes with 'external nodes,' likely TSMC​

Not that long ago, some Intel faithful were in denial that one (of 2) CPU tiles, the high end one would be on TSMC. They believed all would be Intel 20A.

And let's not forget, the 8+32 Arrow Lake on 20A they just "knew" was coming...

Now, Arrow Lake is all TSMC.

But, on the bright side, Intel foundry got a "packaging win", from Intel itself, to package all the TSMC die.

Intel announces cancellation of 20A process node for Arrow Lake, goes with external nodes instead, likely TSMC [Updated]

Out with the Intel node, in with TSMC.

www.tomshardware.com

This is wild actually. Doesn't sound like 20A was anywhere close to N3B for their goals!

 

[gdansk]

I thought that was known already. Or was it still thought some parts would be 20A?

 

[Joe NYC]

I thought that was known already. Or was it still thought some parts would be 20A?

I think the expectation was that the high end die, 8+16 was going to be TSMC and the low end die, 6+8, was going to be Intel 20A.

 

[DavidC1]

3.5 years into new management. Let's see what we got.
-Completely flubbed on demand expectations. Somehow expected the artificial increase to continue
-Cancelled Falcon Shores hybrid until who knows when? Now a GPU only variant is maybe 2025, which AMD/Nvidia has already.
-Intel 3 was supposed to offer density gains at the high performance line too. The actual Intel 3 is only 10% gain, if you chose the "HD" variant with lower performance
-18A was supposed to be 10% over 20A, and 20A was supposed to be 15% over Intel 3. Now we find out 18A is only 15% over Intel 3, with a disappointing 30% density gain. So really rather than 5N4Y, which with extra performance and density was somewhat justified, we have: 7, 4, 4+, 20A, 20A+, or 3N4Y, assuming we count 7, which makes no sense anyways but let's give them that.
-Qualcomm left, Softbank left, and now Broadcomm is on a thin line.
-The whole IDM 2.0 was about separating the two enough, not eventual goal of complete separation of the business. Big fail there. Since Pat also failed demand projections this one is also on him.

 

[desrever]

-Qualcomm left, Softbank left, and now Broadcomm is on a thin line.

Intel left too

 

[Joe NYC]

Intel left too

But Intel Foundry won Intel Products back for packaging of TSMC dies.

 

[DavidC1]

But Intel Foundry won Intel Products back for packaging of TSMC dies.

Intel management: FREE FRUITS AND COFFE BACK ON THE TABLE!!

 

[DAPUNISHER]

Ohio locals are getting nervous - https://www.nbc4i.com/intel-in-ohio...ant-with-possible-changes-to-parent-division/

 

[dttprofessor]

18A d0<0.4 now ,who care 20a?
Intel must make some money through lnl/arl/gnr/gaudi(whatever be produced in anywhere) to feed themself & fabs.
No profit even loss with 50b revenue is unacceptable.

 

[DavidC1]

18A d0<0.4 now ,who care 20a?

Who cares 20A? We do because the current "18A" has 20A specs.

I'm no longer even confident on Clearwater Forest, if they are going to change such important information on a whim.

I wouldn't be surprised if in the USA, where people are trigger-happy will bring another class action lawsuit over lying about 18A.

 

[DrMrLordX]

Intel announces cancellation of 20A process node for Arrow Lake, goes with 'external nodes,' likely TSMC​

Not that long ago, some Intel faithful were in denial that one (of 2) CPU tiles, the high end one would be on TSMC. They believed all would be Intel 20A.

And let's not forget, the 8+32 Arrow Lake on 20A they just "knew" was coming...

Now, Arrow Lake is all TSMC.

But, on the bright side, Intel foundry got a "packaging win", from Intel itself, to package all the TSMC die.

Intel announces cancellation of 20A process node for Arrow Lake, goes with external nodes instead, likely TSMC [Updated]

Out with the Intel node, in with TSMC.

www.tomshardware.com

Whaaaaaaaaaaaaat

Is Intel going to have to tape out the 6+8 tile on N3B, assuming they've already done the work on 20a? Or are they gonna scrap the 6+8 tile or delay it? I'm envisioning a future where Intel can sell you Arrow Lake-U (Meteor Lake, ha ha!) and Lunar Lake but nothing else.

 

[Kepler_L2]

Intel announces cancellation of 20A process node for Arrow Lake, goes with 'external nodes,' likely TSMC​

Not that long ago, some Intel faithful were in denial that one (of 2) CPU tiles, the high end one would be on TSMC. They believed all would be Intel 20A.

And let's not forget, the 8+32 Arrow Lake on 20A they just "knew" was coming...

Now, Arrow Lake is all TSMC.

But, on the bright side, Intel foundry got a "packaging win", from Intel itself, to package all the TSMC die.

Intel announces cancellation of 20A process node for Arrow Lake, goes with external nodes instead, likely TSMC [Updated]

Out with the Intel node, in with TSMC.

www.tomshardware.com

With 18A delay and 20A cancellation, Intel has gone from 5 nodes in 4 years to 4 nodes in 5 years.

 

[Kepler_L2]

Whaaaaaaaaaaaaat

Is Intel going to have to tape out the 6+8 tile on N3B, assuming they've already done the work on 20a? Or are they gonna scrap the 6+8 tile or delay it? I'm envisioning a future where Intel can sell you Arrow Lake-U (Meteor Lake, ha ha!) and Lunar Lake but nothing else.

There were two 6+8 tiles, one on N3B and 20A. What's interesting is that AFAIK there was a small number of ARL samples shipped to OEMs/AIBs using the 20A 6+8 tile, so it was a working node, but perhaps cost/yield was much better on N3B.

 

[poke01]

With 18A delay and 20A cancellation, Intel has gone from 5 nodes in 4 years to 4 nodes in 5 years.

18A got delayed what? Isn’t it supposed to launch next year?

 

[DrMrLordX]

There were two 6+8 tiles, one on N3B and 20A.

Intel had a contingency. Smart. Costly, but smart.

What's interesting is that AFAIK there was a small number of ARL samples shipped to OEMs/AIBs using the 20A 6+8 tile, so it was a working node, but perhaps cost/yield was much better on N3B.

Intel will be wanting those back immediately. Anyone who can come up with a board featuring one of these canceled 20a parts will have an interesting piece of history.

 

[TESKATLIPOKA]

Doesn't look good for Intel, lowest valuation in 15 years, If I include inflation then likely the lowest in decades.
I think It's very unlikely they can keep the foundry business.
If they go fabless, they will certainly survive, but the foundry may not survive unless the government pours billions to keep It afloat. Considering how big the US government deficit is each year, this won't be an easy decision to support.

BTW, Germany paid billions to Intel to build a plant in Germany, but what will they even make? Interposers?