AnandTech recently interviewed Intel's (NASDAQ:INTC) Aicha Evans. She is VP of the company's "Communication and Devices Group." According to her profile on Intel's website, she's in charge of "driving wireless engineering for multi-comm products and Intel platforms." Think cellular modems, NFC, Wi-Fi, GPS, and RF.
Although I highly recommend reading the entire (lengthy) interview when possible, I'd like to point out the three most-important takeaways that I got from the interview.
The "barrier to volume"
AnandTech's Ian Cuttress asked Evans the following question with respect to Intel's mobile chips: "In terms of barriers to volume, is it a case of building the right product, or is there something missing from the portfolio?"
Evans said that "it's all to do with the incumbents." She indicated that the competition -- think MediaTek and Qualcomm (NASDAQ:QCOM) -- is "very, very good," and entrenched. Interestingly enough, Evans was candid in saying that the company's wireless products today are "good," but "don't deliver anything extraordinary and disruptive."
Her rationale for why customers might be interested in using Intel parts that are competent-but-nothing-special was that once Intel moves its phone chips inside of its own factories onto its own manufacturing technology, it will be "able to do things that others cannot easily do."
She suggested that there is a lot of collaboration that needs to happen between the chip vendor and device maker. By starting that collaboration early with Intel's current products, the implication is that partnering on future products based on more-competitive future Intel chips will be easier.
Mobile requires flexibility
Cuttress noted that Intel's initial Atom x3 processors feature ARM (NASDAQ:ARMH) graphics alongside Intel's own CPU cores. He then asked if, with the 14-nanometer Atom x3 parts, Intel would drop ARM graphics, and go with its own internal graphics solution (currently found on the Atom x5/x7 chips). The answer was quite interesting.
"Not necessarily. In mobile you have to learn not to be religious," she said in the interview. "So in mobile you have to really learn to be ambidextrous, so I will use the [intellectual property] that is suited for the appropriate segment."
She also talked about research-and-development efficiency. Evans suggested that, if something is available off the shelf, and if using the ready-made technology doesn't represent a "strategic threat," then it might make sense to keep using technology from a "frenemy" like ARM.
Bringing its A-game to Atom x3
Cuttress asked Evans about the next step for Intel's Atom x3 parts. Evans said that next on the roadmap is to bring these parts to Intel's 14-nanometer technology. She also said that when/if the company can actually deliver Atom x3 parts on that manufacturing technology, "so many new possibilities" open up to the company.
Interestingly, Evans said that Intel needs to be "courageous and take some of this innovation we are driving and have the courage to bring it into a SoFIA class device."
I take this to mean that Intel needs to be willing to be aggressive on features/technologies even in low-end chips. This might not be great for gross profit margins per unit -- more features mean higher costs -- but if Intel can deliver a competitive part that sells well at lower gross margins per unit, that's a lot better than being stingy on features/die size, and selling practically nothing.