Qualcomm: COO Mollenkopf Sees Edge on Intel in Tablet Chips
By Tiernan Ray
Following Qualcomm‘s (QCOM) analyst day meeting in New York today, president and COO Steve Mollenkopf was kind enough to talk with me for a few minutes regarding a variety of topics, including emerging market smartphone trends and the battle for tablet computing.
On the latter score, I asked Mollenkopf how he sees Qualcomm faring relative to Intel (INTC), which has received accolades for its latest “Atom” application processor for tablets, “Bay Trail,” released earlier this year.
Mollenkopf assures me that “given the performance we can deliver with Snapdragon 800 or 805, we are not too concerned about where we sit with respect to other vendors,” referring to the company’s most recent models of its application processor family.
On a deeper level, Mollenkopf rejects the notion, propounded by Intel, that tablet computers will be something more like the traditional personal computer, as the category matures in power and performance, supposedly tipping the advantage to Intel as the PC heavyweight. It is a view that was espoused by former Intel CEO Paul Otellini, and that has been taken up by his successor, Brian Krzanich.
“The tablet is not really so much like a PC,” says Mollenkopf.
“There are a number of things that need to be integrated, and here the PC model basically comes up short, such as video.” He offers as an example the Amazon.com (AMZN) “Kindle Fire HDX” tablet released in recent weeks. It uses a Snapdragon 800. “There are many things we did there that differentiate us, things such as the display, to make it more readable in direct sunlight, for example — those are things you wouldn’t do if it weren’t a mobile product.”
“We also incorporated special security features [in the chipset] for the download and storage of rental content on the Kindle Fire.”
In sum, then, the “we look at the tablet as more of a mobile device, more like a phone, and not a laptop.”
When I suggest that Intel has long dominated the use of performance benchmarks to push the prowess of its chips, Mollenkopf bats that aside, replying “As far as benchmarks for tablets, we have created more realistic benchmarks than the stuff that typically gets thrown around in the PC world, benchmarks that are more attuned to the demands of mobile.”
I also asked Mollenkopf about how he sees the next big leap in chip manufacturing playing out, and whether it may prove problematic given that Qualcomm last year suffered through delays in getting parts at the latest technology node, using features measuring 28 billionths of a meter.
Would there be the same problem with this latest technology process, 20-nanometer, I asked?
“20 nano will not be a drama,” he says. “What happened there was, we had so much traction that we couldn’t match the ramp on the front end.”
“But generally speaking, the problem you just had is never the one you will have next,” he says with a wry smile, before adding, more seriously, “We are converting the supply chain from being node-laggard to node-leader, so things will be different this time.”
I ask Qualcomm about Intel’s vaunted prowess in semiconductor process technology, and Intel’s contention, held by some analyst on the Street, that Qualcomm’s partners, such as Taiwan Semiconductor Manufacturing (TSM), may run aground as chip features face more daunting effects at small feature sizes in coming years.
On the one hand, “we have many partners for making chips,” he points out, “and we think there are many companies that will do just fine making transistors at these new technology nodes, including Samsung [Electronics (005930KS), and TSM, and we think GlobalFoundries has a shot as well. We also fab with [China-based] Semiconductor Manufacturing International. So, we think there are a lot of parties who are capable of making transistors.”
At the same time, “We think, unlike our friends up north [Intel], that the transistor is not necessarily the most important thing,” argues Mollenkopf.
---
Refer to comments in the end.
