Arm CEO: Intel Far Behind in Smartphone Market
For more than a decade Arm Holdings has designed chips that have powered mobile handsets and smartphones like Apple's iPhone. The company now faces a challenger in Intel, which recently demonstrated a smartphone, LG Electronics' GW990, based on its upcoming Moorestown platform.
While Intel tries to move downstream into smaller devices, Arm is aiming at Intel's turf with products for tablets, e-book readers and low-cost laptops, which Arm partners refer to as smartbooks. Lenovo recently announced the low-cost Skylight smartbook and Hewlett-Packard said it is also developing one. Both will be powered by Qualcomm's Snapdragon processor, based on an Arm design.
But Arm faces challenges. Intel's chips go into most netbook-type laptops today. And the low-cost laptops based on Arm processors do not support Microsoft's ubiquitous Windows OS, which is designed for Intel's x86-type chips. However, Arm hopes devices based on Google's Linux-based offerings -- Android and Chrome -- will be a hit with users.
Arm CEO Warren East sat down with the IDG News Service at the International Consumer Electronics Show last week to talk about competition with Intel, compatibility with Windows and Arm's future processors.
IDG: Intel showed the first smartphone with a 5-inch screen. Are you feeling Intel's presence in the smartphone market?
East: Of course. We've been saying that the 100 percent share of applications processors in phones that we have ... can't continue. We don't really see Intel making meaningful inroads into it, not for many years, probably never. In order [for device makers] to switch architectures, the Intel product has to be significantly better to outweigh the cost of switching. We do see in a few years time Intel getting closer to parity with Arm. But that's a few years out yet before they even get closer to parity.
IDG: Arm certainly has an established market in smartphones. But Intel has a strong road map and manufacturing capabilities. How does Arm stack up?
East: People talk about Intel's road map and seem to assume that the Arm road map stands still. Well, the Arm road map doesn't stand still. We move on from Arm11 to Cortex-A8 to Cortex-A9 products. We're licensing the generation after Cortex-A9 at the moment with huge levels of performance. Moving on, we have road maps with 64-bit and virtualization.
Then you come to the semiconductor technology and people say 'Well, Intel's got some superior semiconductor technology.' Maybe they are six months ahead of the likes of TSMC, GlobalFoundries and IBM. We taped out our first 22-nanometer structures the other day. We'll have 32-nanometer microprocessors in volume in the middle of this year from some Arm partners. So there isn't a process advantage either. They are going to make some progress, inevitably, because it's a big market and people are going to want to try different things. But I can't see a way for them to make meaningful progress.
IDG: Intel has said that one of its advantages is the software development ecosystem, an area where they imply Arm is weak. How have you been developing the software ecosystem to support your devices?
East: The reality is the Arm ecosystem is really strong. I think when Intel made their comments quite some time ago, what they were really talking about is the PC software ecosystem around Arm. Absolutely, it's true -- certainly it was true -- that PC applications have been targeted for the last 25 years at the Intel processor, not at Arm. But in this world of the mobile Internet going forward, no, that's Arm.
To make a good consumer experience, you need good browser experience, good plug-ins, good operating systems. These are the leaders in the operating system world of consumer devices. Are they the leaders in the operating system world of PCs? Of course not. PCs have got 25 years of Microsoft history behind them and Windows is what you expect if you want a grey box that looks like a [laptop]. A consumer doesn't want to know about the operating system... [or] of the technology getting in the way. There's a whole load of browsers, plug-ins and enablers [for Arm].
IDG: Arm said it would focus on application processors for servers but no products are out yet. Is that a major market going forward for the company?
East: To say "progress" is a little bit strong. But servers are an emerging application for us in a few years at the back-end of the cloud. Are we going to be in supercomputers for space applications and weather forecasting? Forget it. But the back-end of the cloud, absolutely. Power consumption and cost are key issues for those people, and Arm is the way forward. We've been working for about two to three years with some startups, but now with some of our mainstream semiconductor partners ... and some blue-chip system companies who produce servers and equipment.
IDG: Has the emergence of low-cost laptops like netbooks and smartbooks taken you by surprise?
East: Smartbooks and netbooks ... we sort of expected it. Guessing the form factor is something we don't spend a lot of time on. Our goal is to enable semiconductors, to enable the equipment makers to ... capture consumers. We think our technology helps because of the very low power consumption. It also enables very high levels of integration, that means low-cost semiconductors. Apple, when they did the iPhone, they really moved the consumer experience on it, made it easy and took the technology out of the way. What we're doing is providing the engines to enable people to do that.
IDG: Microsoft has mentioned that its operating systems like Windows 7 won't support Arm. Are you pushing Microsoft for Arm compatibility with Windows?
East: I'm confident in our world, with or without Microsoft. If Microsoft wants to play, then good because they will realize some opportunities that they would otherwise lose. We would think that either there will be large Windows support from Microsoft, or Microsoft would have missed out on an opportunity. That's a decision only Microsoft can make, we can't do it for them. Would we accelerate the Arm progress for form factors with the Windows logo? Yes. Any improvement on that is good for us. If Windows doesn't happen, we can manage without the acceleration. I can sympathize with Microsoft because it's a difficult challenge for them because they have 25 years of baggage behind [PCs].
IDG: What is that baggage?
East: The issue is the old printer that you kept under the spare bed. When you bought a new PC and ... you plugged in your old printer with Windows, it just works. But they have to make all that baggage available on Arm as well. So it's a big maintenance task.
IDG: There are still hardware compatibility and usability issues in Linux that will take time to resolve. That could be an issue for people used to Windows.
East: If you look at what's been accomplished in 18 months to two years and compare that to the 25 years of the Wintel monopoly, it's clear that the consumer experience on Linux-based desktops ... in a relatively short period of time is catching up quickly. It's not going to be an issue.
IDG: What will Arm put in its next designs?
East: What's happening in mainframe land is happening in integrated chips. You can certainly extend it to 64-bit, virtualization and so on. Those are the type of things you will see. Whenever we're doing a new microprocessor, we're always looking at more miles per gallon. As we shrink the silicon you get the speed up as well. We do classic computer architecture ... it's all about getting more mathematics done in the same period of time.