Intel CEO Says Using Competitor's Semiconductor Process in Intel Fabs is an Option

[AleksandarK]

Semiconductor manufacturing is not an easy feat to achieve. Especially if you are constantly chasing the smaller and smaller node. Intel knows this the best. The company has had a smooth transition from other nodes to the smaller ones until the 10 nm node came up. It has brought Intel years of additional delay and tons of cost improving the yields of a node that was seeming broken. Yesterday the company announced the new Tiger Lake-H processors for laptops that are built using the 10 nm process, however, we are questioning whatever Intel can keep up with the semiconductor industry and deliver the newest nodes on time, and with ease. During an interview with Intel's CEO Bob Swan, we can get a glimpse of Intel's plans for the future of semiconductors at the company.

In the interview, Mr. Swan has spoken about the technical side of Intel and how the company plans to utilize its Fabs. The first question everyone was wondering was about the state of 10 nm. The node is doing well as three Fabs are ramping up capacity every day, and more products are expected to arrive on that node. Mr. Swan has also talked about outsourcing chip production, to which he responded by outlining the advantage Intel has with its Fabs. He said that outsourcing is what is giving us shortages like AMD and NVIDIA experience, and Intel had much less problems. Additionally, Mr. Swan was asked about the feasibility of new node development. To that, he responded that there is a possibility that Intel could license its competitor's node and produce it in their Fabs.



Just like GlobalFoundries licensed 14 nm technology from Samsung Electronics to produce the node at GlobalFoundries facilities, the same would apply to Intel. However, that is only considered as an option for now. Intel has made a lot of development on the next-generation nodes as well, and that is not counting the current 10 nm one. The 7 nm is going to arrive shortly, and even smaller nodes are in R&D phases. Licensing a node from someone else would simply null Intel's efforts so we have to wait and see how it plays out. Simply put, Intel has no plans in dropping its Fabs as it is the company's competitive advantage and it only plans to grow.

View at TechPowerUp Main Site

 

[TheLostSwede]

Hang on, is Intel actually admitting they have competition now?
And why would this competition allow Intel to use their process in Intel's fabs?

 

[dj-electric]

Intel has to be the hardest nut to crack in the tech world, I swear.
Their high horses are so high, they will probably mock their competition while bleeding to death with several bullet wounds in their chest. You can see it in every high position adress to investors, in every PR slide.

Swan's idea is not as crazy as some of us might think, and personally I would prefer silicon being made in USA, Israel, and other fabs and not just centered in Taiwan. How feasable it is? I think that money should do the talking.

 

[Vya Domus]

Of course it is, they can't arrogantly bury themselves forever with manufacturing that doesn't work out.

Just one problem, the pie isn't big enough for everyone.

 

[londiste]

Three fabs at 10nm?
Doesn't Intel have like 10-15 fabs in total? I thought at least half of them are on 10/7nm track.

And why would this competition allow Intel to use their process in Intel's fabs?

Lots and lots of moneys?

Just one problem, the pie isn't big enough for everyone.

At this point, with increasing prices and apparently increasing shortages when only 3 manufacturers are left in the cutting edge space? There is enough pie for a while. Licensing something like 7nm today would probably not hit TSMC too hard. From what we know of Samsung's foundry business, they are mostly doing in-house stuff similar to Intel so they would not care much or at all.

 

[DeathtoGnomes]

Hang on, is Intel actually admitting they have competition now?
And why would this competition allow Intel to use their process in Intel's fabs?

Intel got pressure from a hedge fund so now they are looking at options.

 

[Patr!ck]

As long as TSMC or Samsung produce my Xe-HPG gaming GPUs, I'm fine with it.

 

[Frank_100]

Hang on, is Intel actually admitting they have competition now?
And why would this competition allow Intel to use their process in Intel's fabs?

Royalties,. Everyone loves royalties.

 

[TheLostSwede]

Intel got pressure from a hedge fund so now they are looking at options.

Royalties,. Everyone loves royalties.

Sure, but that wasn't my question. The question was, WHY would Samsung or TSMC allow Intel to use their process node in an Intel factory? That would pretty much be giving away your technical lead to the competition, which no sensible company would do.

 

[mtcn77]

Their high horses are so high, they will probably mock their competition while bleeding to death with several bullet wounds in their chest.

I don't think this is fair. You take a working company, chop its veteran workforce, and suddenly how come Intel isn't competitive anymore?
The truth is, it takes two to tango and Intel execs try to do it on their own. That is not very probable - you may waltz alone, but tango? I don't think so...
Intel needs not forget what a beacon it once was. They wouldn't even let their company be compared with AMD. It accompanied pride. Suddenly, mediocre people are rising in the ranks.
I have a friend in the industry - who happened to be an ardent Intel fan with goals of making himself an i9 Starcraft inspired pc, he was a competitive Starcraft player - he didn't get a job in Intel. I don't know the details, but something is not right when some very bright and capable 'fans' cannot catch their dream job at Intel.
Something is very wrong at that company and I hope for the best, for Intel isn't just another player in the PC market. They are all that they said they were - they are the first to the finish line.

 

[dj-electric]

who happened to be an ardent Intel fan with goals of making himself an i9 Starcraft inspired pc, he was a competitive Starcraft player - he didn't get a job in Intel

Being a fan is one thing, working there is another. No matter how much of an advocate you want to be for one brand or another, it takes experience, education and capability to work at places like Intel, and not "being a fan and create a Startcraft i9 custom PC". Silicon engineering and project development is a serious business.

You say "chop its veteran workforce". Where is that from? I know Intel, I know the people who work there, and had chats with some of their top engineers including those who are responsible for development of recent and upcoming products. Still do, btw, on half-year basis. As a part of me belonging to a media body.

These are nothing short of stone cold veterans, who have been working at Intel for decades.
The rest is completely out of context.

 

[Xaled]

Copying the Samsung-Apple experience !? Didn't Samsung learn how to make good processors only after they manufactured Apple's processors?

 

[iO]

Sure, but that wasn't my question. The question was, WHY would Samsung or TSMC allow Intel to use their process node in an Intel factory? That would pretty much be giving away your technical lead to the competition, which no sensible company would do.

They don't really compete with each other, Intel is only making stuff for themselves as their foundry business is dead. And a license would very likely include a clause which forbids Intel to offer any foundry business based on Samsungs/TSMCs node.

But it sounds more like something to soothe investors like "we have all the options, all will be fine" rather than a realistic possibility...

 

[mtcn77]

it takes experience, education and capability to work at places like Intel, and not

No rhetoric, please. He got the job at AMD 3 promotions until he is a fellow. I think that suffices his EE qualifications.

 

[TheoneandonlyMrK]

No rhetoric, please. He got the job at AMD 3 promotions until he is a fellow. I think that suffices his EE qualifications.

Fair enough , probably should have run with that as his CV instead of him building an i9 StarCraft pc being enough.
For anyone offering work above £1000000 I built 100s of PC's that could play Crysis..

 

[mtcn77]

Fair enough , probably should have run with that as his CV instead of him building an i9 StarCraft pc being enough.
For anyone offering work above £1000000 I built 100s of PC's that could play Crysis..

Like I said, this is one of those things in life that doesn't follow logic in all truthfulness. I gave that example just to show his leanings.

These are nothing short of stone cold veterans, who have been working at Intel for decades.

So is he. He is great at what he does, you cannot get a read on him. He works round the clock. While we were training on the usual curriculum, he would study from the old manuals. I think Intel executives should know what they are missing out.

 

[DeathtoGnomes]

But it sounds more like something to soothe investors like "we have all the options, all will be fine" rather than a realistic possibility...

Sure, but that wasn't my question. The question was, WHY would Samsung or TSMC allow Intel to use their process node in an Intel factory? That would pretty much be giving away your technical lead to the competition, which no sensible company would do.

I think the WHY is due to shareholder pressure, thats why i mentioned the hedge fund story, I do agree it doesnt seem sensible

 

[TheLostSwede]

I think the WHY is due to shareholder pressure, thats why i mentioned the hedge fund story, I do agree it doesnt seem sensible

Sensible from a shareholder point of view, maybe. Not so much from a competitor point of view.

 

[mtcn77]

Sure, but that wasn't my question. The question was, WHY would Samsung or TSMC allow Intel to use their process node in an Intel factory? That would pretty much be giving away your technical lead to the competition, which no sensible company would do.

Yeah, it is so absurd it is against the set precedent. Why would Intel be encumbered with fabrication facilities in house during all this time, if they could have outsourced it from a 'fellow' competitor anyway?
It is like one of those unenlightened hasty conclusions that come to mind if you don't think it over. It is a nonsequitur assertive fallacy.
Intel does not collaborate. Intel buys a company it so needs. Industry collaboration cancels your market dominance.

 

[TheLostSwede]

Yeah, it is so absurd it is against the set precedent. Why would Intel be encumbered with fabrication facilities in house during all this time, if they could have outsourced it from a 'fellow' competitor anyway?
It is like one of those unenlightened hasty conclusions that come to mind if you don't think it over. It is a nonsequitur assertive fallacy.
Intel does not collaborate. Intel buys a company it so needs. Industry collaboration cancels your market dominance.

It seems like they might be outsourcing the production of some of their niche computing GPUs though, if rumours are true.
Still very different from having TSMC or Samsung bring their node tech into an Intel fab.

 

[londiste]

There are too many rumors going around about Intel, TSMC and manufacturing. Probably best to wait for the dust to settle with conclusions.
- Intel will definitely outsource manufacturing to TSMC, GPU manufacturing outside their own fabs is known and TSMC is all but confirmed to get that.
- Intel licensing process node or parts of it for their factories seems possible enough but this might just be a statement to appease shareholders who grumbled around this topic recently.
- Manufacturing CPUs at TSMC is effectively a rumor for now. Possible but still unlikely.

Sensible from a shareholder point of view, maybe. Not so much from a competitor point of view.

Are Samsung and TSMC direct enough competitors?
- Intel does not have foundry business, which is what TSMC is in its entirety. TSMC customers obviously compete with Intel but that is not direct from TMSC's point of view.
- Major parts of Samsung manufacturing is in-house stuff, foundry business is a minor part of what their foundries do. Even then, major customers like SoC partners and Nvidia are not competing with Intel CPUs which is what Intel is mainly manufacturing in their own fabs in foreseeable future.

Of course, coffers of money always help with these considerations

 

[mtcn77]

Still very different from having TSMC or Samsung bring their node tech into an Intel fab.

This is reflecting some doubt into Intel's process which is likely to continue if Intel doesn't lean on its own volume. It shows numbers are stagnant and I mean the process yield numbers. It is striking: every node is built on top of another, should Intel slow down now, who knows how it will affect Intel later on...
PS: I mean Intel will instead be improving its competitors' foundry business if they don't lean on their own volume.

 

[TheLostSwede]

Are Samsung and TSMC direct enough competitors?
- Intel does not have foundry business, which is what TSMC is in its entirety. TSMC customers obviously compete with Intel but that is not direct from TMSC's point of view.
- Major parts of Samsung manufacturing is in-house stuff, foundry business is a minor part of what their foundries do. Even then, major customers like SoC partners and Nvidia are not competing with Intel CPUs which is what Intel is mainly manufacturing in their own fabs in foreseeable future.

Of course, coffers of money always help with these considerations

I meant as a chip maker, regardless of whom the customer is.
Both are competitors as far as chip manufacturing is concerned, as they're the only two other foundries Intel could cooperate with where they could gain an advantage over their own, current technology.
This not meant in the means of competitor as a foundry business, sorry if that wasn't clear.
Intel does actually have some very minute foundry business. There's also the "custom" Apple parts and the 5G modems for Apple that they're still contractually obliged to provide.

This is reflecting some doubt into Intel's process which is likely to continue if Intel doesn't lean on its own volume. It shows numbers are stagnant and I mean the process yield numbers. It is striking: every node is built on top of another, should Intel slow down now, who knows how it will affect Intel later on...

Well, that's normally how it is, but Intel stepped away from all their know-how with 10nm and tried something new. It failed and kept failing. Their "new" 10nm node is as far as I understand, at least in part, based on their previous node. It doesn't seem to be going great so far though, but maybe they'll fix it this year...
Regardless, until they come up with a working node, be it 10nm or 7nm, it seems like they are going to outsource some parts to be able to keep up.

 

[mtcn77]

Their "new" 10nm node is as far as I understand, at least in part, based on their previous node. It doesn't seem to be going great so far though, but maybe they'll fix it this year...
Regardless, until they come up with a working node, be it 10nm or 7nm, it seems like they are going to outsource some parts to be able to keep up.

That is what's so appalling. It actually reminds of AMD's foundry business, GlobalFoundries, having to sign an agreement with Samsung just to continue production in 14 nm and then cancelling 7nm development and AMD having to do something about it.
You just lose sight as a foundry if you cannot scale up from past production. It is more important than anything in the numbers game. You just have to keep the machines running to fine tune your competitive edge.

 

[TheLostSwede]

That is what's so appalling. It actually reminds of AMD's foundry business, GlobalFoundries, having to sign an agreement with Samsung just to continue production in 14 nm and then cancelling 7nm development and AMD having to do something about it.
You just lose sight as a foundry if you cannot scale up from past production. It is more important than anything in the numbers game. You just have to keep the machines running to fine tune your competitive edge.

Depends in what market you're trying to compete. Loads of weird niche stuff that's still at 250 or 180nm, because going smaller doesn't really work.
But yeah, if you want to be a leading edge foundry, you better be focused at what you're doing and not take any big chances, as if you make one misstep, you're going to lose a lot of business.
This is also why the move to EUV is so scary, as until it has been proven in full production, it's a big risk for all parties involved.