Intel Considers Sale of Altera Business Amid Restructuring Plans, Foundry Business to Stay

[AleksandarK]

Intel is reportedly exploring the sale of its Altera business, a move guided by CEO Pat Gelsinger as part of broader restructuring efforts. Acquired by Intel in 2015 for $16.7 billion, Altera, formerly known as the Programmable Solutions Group, has been a profitable segment. However, with Intel facing financial strain due to extensive spending, the company is now considering divesting its FPGA business to recoup capital. Currently, Altera operates as a separate entity within Intel, relying on the tech giant for R&D, sales, marketing, and support. Gelsinger is expected to propose the sale at a board meeting scheduled for mid-September, where he will outline his vision for Intel's future. This restructuring could also affect other parts of Intel's operations, including its Foundry business.

While previous reports suggested that Intel might spin off its Foundry unit or sell it to industry leaders like TSMC or Samsung, the latest information indicates that Intel plans to retain this division, albeit with scaled-back expansion efforts. The $32 billion factory in Germany, for example, may be scrapped, along with other capital-intensive projects, and other capital expansions may also be put on hold. Pat Gelsinger's vision still needs to be finalized and is still in the drafting phase, so until the mid-September board meeting, we have to wait to gain more information.



View at TechPowerUp Main Site | Source

 

[Daven]

“Pat Gelsinger's vision…”

You mean Pat Gelsinger’s failure.

The word vision usually involves positive thinking for the future of a company. What’s happening are plans being made to reduce Intel due to business failures.

 

[R0H1T]

They should've seen this coming, no way those $100 billion worth of "expansions" gonna bear fruit inside 5 years! Did they miss the capacity expansion by TSMC, Sammy & of course China

 

[Ferrum Master]

Poor R&D people.

 

[persondb]

Honestly speaking, Altera is one of the best choice to spin off. Intel didn't really do much to it with the rest of Intel had. They could have done a lot of things like putting x86 cores in the FPGA SoCs (like the ARM ones) or have small FPGAs into their own client/server SoCs and etc.

Unlike AMD which got Xillinx in big part due to their attractive AI IP(Ryzen AI is essentially IP from Xillinx), they really do not have much of a need for Altera. The chips that were designed with Intel processes will continue to being made anyway, because FPGAs need to have that extended support, so they will have the foundry business aspect even if it spins off.

Altera is kinda attractive for AI since FPGAs are good for edge AI, so that at least will be worth quite a bit.

“Pat Gelsinger's vision…”

You mean Pat Gelsinger’s failure.

The word vision usually involves positive thinking for the future of a company. What’s happening are plans being made to reduce Intel due to business failures.

He took over about 3 years ago? This is barely enough time to change a lot of things, specally in a gigant corporation like Intel. Most of the current projects being delivered were probably started under previous administrations.

 

[ViperXZ]

You mean Pat Gelsinger’s failure.

Not really, I’m not a fan of the guy but he’s not at fault, Intel had purely money driven managers for a long time before him and they ruined Intel, Pat came back to repair it and a few years ain’t enough to do that. And according to some people those money driven people aren’t even fully gone from Intel, just not CEOs anymore.

 

[OSdevr]

Not a bad move. Xilinx/AMD dominates the FPGA market and Altera hasn't been very innovative lately.

What's more the free version of Altera's Quartus software has severe limitations on what FPGAs can be used with it compared to Xilinx's which they refuse to budge on. That means only large companies can develop stuff using anything but their lowest end FPGAs (the latest ones with transceivers are a decade old).

 

[thesmokingman]

This is one way to raise your share price.

 

[Darmok N Jalad]

I'm honestly surprised at how bad things have gotten for Intel in such a short time. I suppose the PC boom that took place during lockdown kept their execution failures at bay (mostly fab-related issues), but now that all that has passed, they are really looking ragged. I just didn't think they'd ever end up in a place where they are looking to cut 10B and sell a bunch of assets. I suppose maybe they just got so big that they lost focus on their core business. Still wondering if the GPU division is going to get shut down.

 

[thesmokingman]

I'm honestly surprised at how bad things have gotten for Intel in such a short time. I suppose the PC boom that took place during lockdown kept their execution failures at bay (mostly fab-related issues), but now that all that has passed, they are really looking ragged. I just didn't think they'd ever end up in a place where they are looking to cut 10B and sell a bunch of assets. I suppose maybe they just got so big that they lost focus on their core business. Still wondering if the GPU division is going to get shut down.

Nah, this is what it looks like to see the fall of a once too big to fail giant. It's gradual... but once their cash reserves shrink, that's when they finally start to feel it.

 

[Vayra86]

“Pat Gelsinger's vision…”

You mean Pat Gelsinger’s failure.

The word vision usually involves positive thinking for the future of a company. What’s happening are plans being made to reduce Intel due to business failures.

The man only has a vision of old socks. Him and Raja are like a match made in heaven. Called it years ago

I'm honestly surprised at how bad things have gotten for Intel in such a short time. I suppose the PC boom that took place during lockdown kept their execution failures at bay (mostly fab-related issues), but now that all that has passed, they are really looking ragged. I just didn't think they'd ever end up in a place where they are looking to cut 10B and sell a bunch of assets. I suppose maybe they just got so big that they lost focus on their core business. Still wondering if the GPU division is going to get shut down.

Yeah... IMHO it would be one of the worst possible choices. Parallelization is ever more a thing and CPUs will never surpass GPUs in this type of task.

If you think of it, Intel really tried hard at making many core consumer CPUs but the reality is, I think, that they already knew that going past quad cores would be the beginning of the end for their Core architecture. It could not scale that well for that much. 6c12t was already trouble - delid required to really make them tick well - and it didn't last very long before they got forced to get much better bins generally used for 'Extreme' or Intel E- generations and server chips into the consumer stack. And even then, they had to increase TDP. Eventually they had to scale down the core and present an alternative to keep scaling. And now we're effectively back at a lot of consumer chips with '4 P cores' and a shitload of helpers that while efficient are lacking in instruction sets.

Perhaps AMD's Bulldozer is Intel's Big Little... minus the court case.. oh wait

 

[InVasMani]

Intel's missteps to get to this point aren't small. They look like Wile E. Coyote dropping from a cliff-side right now.

 

[remixedcat]

Oh if they cancel the Ohio plant they will be boned in chickens! They uprooted tons of ppl for that as well as farms. Ohio gon go ham!!

And ohio really needs that fab too!!

If they cancel they should pay the govt back for all the chips act money!!

 

[trsttte]

Oh if they cancel the Ohio plant they will be boned in chickens! They uprooted tons of ppl for that as well as farms. Ohio gon go ham!!

And ohio really needs that fab too!!

If they cancel they should pay the govt back for all the chips act money!!

They'll cancel Magdeburg (Germany) before dropping anything in the US, it's a strategic industry for the US and they wouldn't let it happen any other way

 

[InVasMani]

Cart before the horse. Offloading Altera probably is the right call to make for Intel given their options. The real issue is that it got to this point in the first place. The amount of mismanagement to make it to this point is appalling. This company was run into the ground basically. They left the company on cruise control for too long and are paying for it now. Both TSMC and AMD their two biggest competitors are stronger than ever because Intel got over complacent and played itself. They made Nvidia a monopoly basically too or at least the easiest company to point to on the planet that's as legal of a monopoly as you can get away with. Nvidia really reaped the rewards of Intel monopoly behavior against AMD and not a small part of why they've grown overly dominant today. That of course is my observational view on how it's unfolded.

 

[trsttte]

They left the company on cruise control for too long and are paying for it now.

Not cruise control, the managers at the time were all happy only doing stock buybacks with cheap money and getting huge bonuses for it. It was going fine until the competition woke up and slapped Intel with a good dose of reality.

 

[InVasMani]

Isn't that kind of the same thing!? The higher up people at Intel were complacent with the monopoly money train they created that lasted about a decade until AMD came back with a all bases loaded hit with Ryzen and followed it up up a few times over then hit a grand slam with X3D that's been somewhat of a nail in the coffin for Intel. AMD's pretty said no Intel you can lay rest for decade until you've learned your lesson the hard way. Intel's really dug it's own burial hole ironically though between 10nm screw ups and defect fiasco alongside many other surrounding controversies. There is no one to blame, but itself for Intel's current situation.

 

[Dr. Dro]

The man only has a vision of old socks. Him and Raja are like a match made in heaven. Called it years ago

It's been what, 6 years since Raja Koduri left AMD and Radeon is still their worst performing business. Not to mention that Radeon itself (and by this I mean the original Radeon from 2000) had his personal involvement, as well as the most successful of their GPU designs, Polaris. RDNA and CDNA split occurred for a reason, they had been re-heating GCN for 10 years at that point. Not even if you brought Nvidia's most experienced engineers in they could fix GCN by the time Vega 20 came out.

The problem runs much deeper than that, and using him as a scapegoat has grown quite old. The copium has gotta run out someday, sheesh.

 

[persondb]

Offloading Altera probably is the right call to make for Intel given their options. The real issue is that it got to this point in the first place.

I would argue that their issue is that they did literally nothing with Altera aside from forcing Altera to use Intel nodes, which was more of a detriment to Altera vs Xillinx due to Intel issues with their processes. And due to that lack of focus and usage, a lot of Altera was allowed to degrade and lose to Xillinx.

As an example, they don`t have developed as much of the big fuss AI IP for FPGAs as Xillinx does(and which ended up being used for other AMD products).

as well as the most successful of their GPU designs, Polaris

Considering how long those projects take, he probably had some influence on RDNA 1(and by extension RDNA 2 as RDNA 1 was essentially a prototype to it) which was pretty successful(with RDNA 2 being even more so).

But he will forever be remembered by his hype on Vega.

Plus Lisa Su is considered to be a Savior Saint and thus it couldn't possibly be (partly) her fault, so Raja it is.

 

[Neo_Morpheus]

Just as AMD was forced to do, thanks to Intel illegal actions, its now their turn.

But karma alone is doing this and fingers crossed they never recover.

 

[kondamin]

They'll cancel Magdeburg (Germany) before dropping anything in the US, it's a strategic industry for the US and they wouldn't let it happen any other way

That would be an other serious nail in the us/germany coffin

I think the best option would be intrest free loan and having the project go ahead considering the wider geo political consequences.

 

[Chaitanya]

They'll cancel Magdeburg (Germany) before dropping anything in the US, it's a strategic industry for the US and they wouldn't let it happen any other way

If they cancel Germany plant they will have to cough up billions with interest they leeched in subsidies from German govt. If AFD vetos those payments then it would be easy to cull Germany "expansions".

German budget woes threaten chip fab funding

Constitutional court tells govt: Er, about that €60B you handed out... it's not legal

www.theregister.com

 

[utmode]

First pocket Govt grant money then shrink business, citizens are always loser.

 

[Vayra86]

It's been what, 6 years since Raja Koduri left AMD and Radeon is still their worst performing business. Not to mention that Radeon itself (and by this I mean the original Radeon from 2000) had his personal involvement, as well as the most successful of their GPU designs, Polaris. RDNA and CDNA split occurred for a reason, they had been re-heating GCN for 10 years at that point. Not even if you brought Nvidia's most experienced engineers in they could fix GCN by the time Vega 20 came out.

The problem runs much deeper than that, and using him as a scapegoat has grown quite old. The copium has gotta run out someday, sheesh.

You should readup a little on the early history of Xe I think because Raja was at the helm of strategic blunders quite often. This wasnt a story of AMD, its one of Intel failing forward. Just hiring some heads apparently does not a good company make... And the exact same applies to Pat. Plus, what Raja is remembered most by is his PR flair... the man is an elephant in a China shop, just unfit for camera duty.

All I see is old boys networks and they suck harder by the year. Raja is a symptom, as is Pat, of a company and possibly a generation that has lost its touch.

If they cancel Germany plant they will have to cough up billions with interest they leeched in subsidies from German govt. If AFD vetos those payments then it would be easy to cull Germany "expansions".

German budget woes threaten chip fab funding

Constitutional court tells govt: Er, about that €60B you handed out... it's not legal

www.theregister.com

I think that storm has passed already hasnt it? It was mostly pointed at Ukraine support too. Or was the problem moved over to funding fabs now, I might have missedthat.

 

[Chaitanya]

I think that storm has passed already hasnt it? It was mostly pointed at Ukraine support too. Or was the problem moved over to funding fabs now, I might have missedthat.

I was refering to the recent election results from East Germany where a far-right AFD came out as victor and even if it doesnt get to form local governement it has enough numbers to veto bills. So if they veto the funding for new fabs then Intel can basically quit German expansion plans without any repercussions.