Investors following Intel (NASDAQ:INTC) know the company has been making a greater push into the semiconductor foundry business. One of the company's biggest clients to date has been FPGA vendor Altera (UNKNOWN:ALTR.DL). Intel announced its deal with Altera in February 2013, but chips will not begin to sample to customers until the second half of this year. This suggests a material revenue contribution to Intel in 2016.
At the recent Goldman Sachs Technology and Internet Conference, Intel CEO Brian Krzanich explained exactly what this deal with Altera means for his company's custom foundry business as a whole.
What do foundry customers want to see from Intel and Altera?
When asked by Goldman's Jim Covello about Intel's deal with Altera, Krzanich said this:
When people look at Intel from a foundry perspective, they want to see two things. One, can we be a good partner -- we don't have a history of being in the foundry, we don't have a history of being a warm, welcoming company into our factories. This is not what we've done in the past.
So people looking at, can somebody come into our building and work with us? Can they use our design tools, can they get our silicon to work? I think that's the first test that people are waiting to see pass.
The second test is, OK, we see what Intel is able to do with their Moore's Law advantage, can others capitalize on that as well with their architectures? I think that'll be the second test.
Krzanich said he believes Intel will pass both of those tests with Altera. After that, the CEO believes that if all goes as he expects, the doors will open up to "more strategic discussions" with other customers for other foundry deals vis-à-vis custom foundry.
Building chips for competitors? Krzanich says yes
Covello pointed out that much of the foundry business is focused on the mobile market, in which Intel is trying to compete with its Atom processors. He asked Krzanich the following:
Do you view you selling your own mobile chips -- which is preferable because, as you've stated before, you get paid for both the design and the manufacturing -- do you see it as an either/or? Bigger foundry effort which would have to, in some way, be driven by mobile because that's where most of the big foundry customers are or Intel selling its own silicon into the mobile space?
Krzanich's response left little doubt about his take on the issue:
No. That market is big enough, our share is low enough, that I think that the shareholders would prefer that we go out and capture as much of the profit and as much as the revenue and margin dollars that we can.
Krzanich added that the market is large enough that he would "absolutely take competitors' parts" for production at Intel foundries.
Will competitors trust Intel?
Covello then asked whether competitors would feel comfortable having Intel build their parts. Krzanich said first he believes that before most potential customers would be comfortable with this situation, Intel would need to pass the aforementioned set of tests with Altera. "I have to prove to the marketplace that there's a reason they need to be there," he said.
Next, Krzanich said potential customers would "need to have a comfort that [Intel] is not going to disadvantage them." For example, Krzanich explained, customers need to be sure Intel isn't going to release parts that compete with its customers' parts before its customers do.
Krzanich, however, seemed optimistic that those issues could be worked through.