You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Some interesting data about Intel's foundry business emerged at the Citi conference. It seems that they are charging a premium for their foundry services, in order to maintain the high margins that they are accustomed to. Also, they do not expect business to grow very fast:
And what Stacy has said throughout our calls is we believe we can grow that foundry business at a reasonable rate with not impacting our margin structure. And that’s because when people see these cost advantages in terms of Moore’s Law they can afford to pay their margins that we’re talking about. So, we’re really looking for customers like that. It will grow at a rate that it’s not going to be an extremely fast or large part of our business over the next two or three years, it takes - once you even acquire a customer it takes them two to three years to get their first product out. So I think it’s going to be - this is a long term play that will play out over the next five to seven years as people start to see how we keep margin down Moore’s Law and we think our leadership in Moore’s Law continues to grow.
At 14nm Intel Custom Foundry has failed, hopefully we can all agree on that now. The handful of customers Intel does have at 14nm will not result in measurable wafer revenue for the likes of Intel. Intel was looking for a big SoC customer and they just did not get one. Also, according to what I have read recently, the famed Intel density advantage at 14nm is in serious doubt as well:
At 10nm Intel Custom Foundry has another chance. Right now Intel is courting Apple and others promising a two year process lead over TSMC and Samsung. IF that is true AND Intel can make peace with ARM, Apple can switch to Intel, no problem. Making peace with ARM is an absolute must however. ARM works very closely with TSMC and Samsung to tune the processes for ARM IP. They do test chips and other close collaborations from start to finish. Without that Apple will have a difficult time getting a leading edge ARM based SoC into the market on schedule and on spec. And if you think Apple is switching to Intel Atom you ate way too many Twinkies as a kid.
Apple's SoC is their own design, not ARM's. They only license the instruction set. Besides that I think most of Intel's foundry customers are using ARM already.
Apple's SoC is their own design, not ARM's. They only license the instruction set. Besides that I think most of Intel's foundry customers are using ARM already.
Apple uses other ARM IP as well as the instruction set. All of which are part of the test chip process I mentioned above. did you know that Apple was also one of the first investors in ARM? It's all in the book:
"At 14nm Intel Custom Foundry has failed, hopefully we can all agree on that now."
I'm no fan of Intel, but I've yet to see evidence that it "failed" given early it is for a foundry service on an immature process.. Seems like more anti-Intel snarking.
Margin structure seems to be the biggest obstacle for the giant to grow further. Atom is not scalable.
If someone could somehow persuade the giant to spin off a foundry business with more flexible margin structure, TSM would be in big trouble.
"At 14nm Intel Custom Foundry has failed, hopefully we can all agree on that now."
I'm no fan of Intel, but I've yet to see evidence that it "failed" given early it is for a foundry service on an immature process.. Seems like more anti-Intel snarking.
Yes I guess that is subjective. How about "has not met expectations". They did close a fab in Arizona which they had hoped the foundry business would fill. I know of many fabless companies here in Silicon Valley that rejected Intel's 14nm so I have a different perspective. I also remember when they launched the foundry business by trash talking the fabless semiconductor ecosystem:
“Being an integrated device manufacturer really helps us solve the problems dealing with devices this small and complex,” Bohr said “the foundries and fabless companies won’t be able to follow where Intel is going.”
“Bohr claims TSMC’s recent announcement it will serve just one flavor of 20 nm process technology is an admission of failure. The Taiwan fab giant apparently cannot make at its next major node the kind of 3-D transistors needed mitigate leakage current, Bohr said.” “Qualcomm won’t be able to use that [20 nm] process,” Bohr told me in an impromptu discussion at yesterday’s press event where Intel announced its Ivy Bridge CPUs made in its tri-gate 22 nm process. “The foundry model is collapsing,”
So yes, I stand by my opinion that Intel 14nm is a foundry fail and I will ad that Mark Bohr's statements are also a FAIL in regards to the fabless semiconductor ecosystem.D.A.N.
It's kind of risky and strange to me. Intel might have already decided that they don't have to "make peace with ARM". Instead Intel thinks they just need to ask the industry to embrace x86 architecture!? Or Intel can't make up their mind of what exactly the Intel's foundries' customers mean to Intel?
