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

[londiste]

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.

When it comes to business, chip manufacturing has a bunch of different segments. In-house manufacturing and foundry outsourcing are pretty separate major things. They are not direct competitors for business.
Besides, TSMC is not really a chipmaker in the same sense as Intel or Samsung.

Intel has money and probably more than a few patents or technologies to share.

 

[TheLostSwede]

When it comes to business, chip manufacturing has a bunch of different segments. In-house manufacturing and foundry outsourcing are pretty separate major things. They are not direct competitors for business.
Besides, TSMC is not really a chipmaker in the same sense as Intel or Samsung.

Intel has money and probably more than a few patents or technologies to share.

I think you are deliberately trying to misread what I'm trying to say.
Chip, as in IC maker, not a final product.
Yes, they aren't direct competitors today, but they are still in the same kind of business, otherwise we wouldn't be having this conversation in the first place.
It's not as if we're comparing Intel with a sausage maker here, is it?

No doubt Intel sits on a lot of money and they're still making a lot of money. However, if they can't move forward soon, the competition is going to overtake them (and the competition in this case would be AMD and all the ARM licensees and what not) and then what? Obviously AMD made a comeback from the brink of extinction, so nothing is impossible. I'm sure Intel still has a lot of good people working there, but something is clearly rotten inside the company or they wouldn't have been stuck at the same place for so long.

 

[londiste]

If TSMC and Samsung were not making ICs using their processes in Intel fabs would not make much sense
All I am saying is that since they are not directly competing, some form of knowledge sharing/selling between them would not be surprising.

When we talk about manufacturing, it is not AMD or ARM licensees that overtake them. It can only be Samsung or TSMC. If it comes to that you can bet Intel's money is as good as someone else's. With less competition in fab business I am quite certain we would not like that outcome.

I also have a feeling that a set of technical problems are excessively blamed on the management here. Sure, management can make it worse in various ways but the root problem is Intel's 10nm (and probably 7nm) manufacturing processes not being viable.

 

[mtcn77]

Sure, management can make it worse in various ways but the root problem is Intel's 10nm (and probably 7nm) manufacturing processes not being viable.

If you are so worried about what is wrong, fine:
They ran out of ideas at 14nm. I can even cherrypick the timescale where they said it. It was present since 2012.
I tried telling it to other people, including editors like Joel Hruska. Nobody believed, nor showed enough patience to check out the source material.
It is the first three lectures:

https://m.youtube.com/playlist?list=PLwYzDeSMOcCwjs0D7I9V6T01QVcJ-1Dho

You might want to stick around until 21:45" to see what happened to 10nm.

 

[Frank_100]

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.

Samsung and TSMC would be getting the royalties on every chip INTC builds with their process. It is risk free money on a depreciating intangible asset. In a few nodes the processes will change for everyone. Hell yes they would allow it.

By risk free I mean neither TSMC or Samsung would have to expand their capacity to profit off selling more chips and the sunk cost of developing the process has already been paid.
Intel would bear the risk and cost of adapting the process to Intel foundries.

 

[TheLostSwede]

Samsung and TSMC would be getting the royalties on every chip INTC builds with their process. It is risk free money on a depreciating intangible asset. In a few nodes the processes will change for everyone. Hell yes they would allow it.

By risk free I mean neither TSMC or Samsung would have to expand their capacity to profit off selling more chips and the sunk cost of developing the process has already been paid.
Intel would bear the risk and cost of adapting the process to Intel foundries.

Risk free you say? So you don't think Intel would take a very good look at their processes and then try to improve their own production lines using the competitions advances if their nodes were implemented in Intel fabs? No, not at all, zero risk...

 

[londiste]

Risk free you say? So you don't think Intel would take a very good look at their processes and then try to improve their own production lines using the competitions advances if their nodes were implemented in Intel fabs? No, not at all, zero risk...

There is risk but it doesn't seem very significant.
None of TSMC customers are going to manufacture their stuff at Intel, perhaps with the exception of Intel itself. Same with Samsung.

 

[TheLostSwede]

There is risk but it doesn't seem very significant.
None of TSMC customers are going to manufacture their stuff at Intel, perhaps with the exception of Intel itself. Same with Samsung.

The risk is that Intel figures out what they're doing wrong and then whoever decides to work with them loses the business a year later.
It really doesn't make a lot of sense to me, but hey...

 

[londiste]

The risk is that Intel figures out what they're doing wrong and then whoever decides to work with them loses the business a year later.
It really doesn't make a lot of sense to me, but hey...

This is what I am trying to get at. What business do you mean they will lose a year later?

 

[TheLostSwede]

This is what I am trying to get at. What business do you mean they will lose a year later?

Intel's business? I don't know if we're reading the same stuff here or now, but what other business is being discussed?
If Intel can "spy" on their foundry competitors by doing this, resolve their problems and move forward on a competitive node, what do they need Samsung and TSMC for?

 

[londiste]

Good point. Intel's business they might indeed lose. Possibly.
I do not think they are hoping to get much of Intel's business though. Everyone expects Intel to get their shit together with manufacturing.

 

[Frank_100]

Risk free you say? So you don't think Intel would take a very good look at their processes and then try to improve their own production lines using the competitions advances if their nodes were implemented in Intel fabs? No, not at all, zero risk...

I see what you are saying but you need to consider that the process is a depreciating intangible asset. The EUV equipment is very expensive and acts as a large moat for both Intel and TSMC. The object of the game is to maximize return on assets.

Have fun with this debate. We will all know in a matter of months.

 

[TheLostSwede]

Good point. Intel's business they might indeed lose. Possibly.
I do not think they are hoping to get much of Intel's business though. Everyone expects Intel to get their shit together with manufacturing.

Time will tell if Intel can actually get their shit together. How many years ago did they promise 10nm parts again?

I see what you are saying but you need to consider that the process is a depreciating intangible asset. The EUV equipment is very expensive and acts as a large moat for both Intel and TSMC. The object of the game is to maximize return on assets.

Have fun with this debate. We will all know in a matter of months.

There's obviously going to be a point in time when an older process becomes less valuable, but at the same time, see above.

If something like this was to happen, it also has to be lucrative enough for both parties, which it might not end up being.