Business Tech

Acer recently outpaced Dell as the second largest PC seller in the world. Is No. 1 Hewlett-Packard next on the list?

Acer Chairman J.T. Wang reportedly intends to ship 40 million notebooks in 2010, a number that could help the Taiwanese company challenge and even overtake current champ Hewlett-Packard for the top spot in portable PC sales.

To achieve this goal, Acer is looking to kick up revenue by more than 70 percent over the next three years, said Wang in an interview with Reuters.

Credit: Acer

J.T. Wang

Much of that revenue growth is likely to come from an anticipated 50 percent gain in sales for low-cost Netbook PCs. As the economy improves, Wang expects to sell an ever greater number of low-priced, portable PCs, a segment that has been very kind to the company.

As chairman since 2008, Wang has led Acer as its sales and market share have shot up over the past year. And with revenue expected to hit $18 billion this year, he's already targeting the next milestone.

Doug Fisher

(Credit: Intel)

For Intel, the driving force behind its Moblin software efforts is its main role in life: a chipmaker.

At the Intel Developer Forum in San Francisco last month, the company showed off the latest version of its Linux-based Moblin operating system, designed for smartphones, Netbooks, and Nettops. In most if not all of those cases, the idea is that the device is built around Intel's Atom processor.

Moblin has been slowly catching on in the Netbook arena. Dell, for instance, recently began selling its $299 Mini 10v Notebook with an option for Ubuntu Moblin Remix, and PC makers such as Acer and Asus are also said to be planning Moblin-based Netbooks. Intel is also positioning Moblin 2.1 for all-in-one Nettop PCs, most of which--such as the Asus Eee Top and Dell's Studio One 19--currently run Windows XP.

I recently spoke by phone with Doug Fisher, vice president and general manager at Intel's System Software Division, about the upcoming latest version of Moblin.

Q: Can you tell me what's new in Moblin 2.1 version versus 2.0? What are some of the features and benefits that 2.1 will offer?
Fisher: The way I look at it, 2.1 is obviously an interim improvement over 2.0. So we have the beta release for the Netbooks out there now that people are playing with. We added an element called a Moblin Garage component. It gives the capability to allow developers to start putting applications in this developer store. It's really designed for putting [out] open-source or fully distributable applications or components of applications that developers can use.

Moblin 2.1 is going to be targeted not only for Netbooks but also moving into other form factors like mobile Internet devices or handheld devices, and in-car entertainment--we call that IVI [in-vehicle infotainment]. It's going to add things like 3G support, a connection manager, telephony framework. The input methods are going to be around touch capability and gesture, so we'll be adding those types of support in 2.1.

Version 2.1 is going to be used not only for Netbook and Nettops, but also for handheld form factors. For things like Nettops, we are working on ensuring the capability to support high-resolution screens. We're also looking at incorporating additional social networking support into the platform to continue evolving on the value proposition that Moblin is defined on.

Q: Are there going to be different versions of Moblin for each hardware platform or will it be basically the same operating system with just minor changes for the platform itself?
Fisher: I love when people ask that question because it really helps deliver the message we are trying to get out with this. The whole Moblin infrastructure is a unified stack across all Atom-based devices. All the form factors use the same operating system, offering environments with the same UI framework, the same application framework, the same core apps. So it's really a unified API across all the devices. The benefit of this is that all the optimizations for footprint, boot time, and battery life, all of that carries across all segments that use Atom at the core level. The UI framework and the applications are consistent across all those. Now that's not to say we don't optimize for form factor and input methods and things like that. But the core elements are consistent across all segments. So the idea is you deliver Moblin-compliant operating environments.

We have 17 OSVs [operating-system vendors] that have publicly committed to this, and I think seven to nine, somewhere in that range, have actually made product level commitments. Obviously, we've heard from Novell and Canonical and others. So, yes, it's the same. Really, it gives the OSVs a clear compliance to align behind Moblin, and more importantly the ISVs [independent software vendors] a standard API set to write to in order to ensure their application can be portable across any Moblin-compliant operating environment.

Q: From the consumer point of view, if somebody is running Moblin on a Netbook, a Nettop, and a smartphone, are they going to see a similar UI and similar usability across the different devices, or will that depend more on the manufacturer and what they do with Moblin?
Fisher: One of the big benefits of Moblin is that the UI is customizable. If you're talking about the Nettop and the Netbook space, you are going to see very, very consistent usage criteria for all those devices. So what the "myzone" looks like on a Nettop will be very consistent with the Netbook. It's customizable, so who ever wants to, an OEM or service provider, can create their branding and imaging and add other capabilities. But the idea is when a user participates with a Netbook or Nettop, it's a very familiar engagement, so it's going to be consistent. We also expect in the handheld space that there will be a lot of consistencies across those devices as well. With that said, manufacturers are going to want to really customize, and it's possible to do that within the UI framework.

Q: I want to talk a little bit about Dell because I know they just came out with a version of their Mini 10v Netbook running [Canonical's] Ubuntu Moblin Remix. What does that refer to? Is that just taking Moblin and coming up with their own customized version? Can you explain what Dell is actually running on their new Netbook?
Fisher: At the highest level, Dell is running Moblin. Moblin is a pure open-source project. It's all open source, so what we rely on is an OSV, like Canonical, to deliver a productized version of the Moblin project. And so what you see in Remix is Canonical's release of the Moblin platform. It's a pure Moblin play, 100 percent aligned with the user interface we described, API-compatible. But they add a lot of value in productizing it, adding other elements that the OEM wants, customizing it.

Q: Can you then talk a little bit about Dell and other notebook vendors looking to run Moblin? What expectations does Intel have for this adoption of Moblin in the notebook market?
Fisher: If you look over the past year and a half, there's been a lot of activities by the OEMs using Linux with Moblin capabilities. But what Moblin has brought is the unification of those capabilities into the release of version 2.0. And what you see from Canonical working with Dell is a unified release. It really is a much more substantial commitment from Canonical and all the other OSVs around this. So what Dell has done is they've released their Mini with [Moblin] to the developer community. The expectation is that in combination with what was announced at IDF, the developer store Intel Atom Developer Program, with platforms from Dell, we really think it's going to drive a lot of innovation around Atom-based devices, in particular around the Moblin development environment. And so down the line, we expect to see OEMs delivering consumer devices as well.

Q: When you think of Linux, you traditionally think of it of it as an operating system that has been adopted more by IT people, technophiles, etc. What do you think has held back the adoption of Linux and what can Intel do to further Linux along so it becomes more of a mass market operating system and not just a niche player?
Fisher: I think there are a couple of things. When you talk about the consumer-visible things like Netbooks, clearly Microsoft is doing great work, and they're delivering a substantial portion of the Nettops and the Netbooks. We had three main areas that we had been driving requirements around, which is Internet usage, media consumption, and social networking. Those are really the three design points Moblin is focused around--that type of device. It's really a different usage experience. What you described is really true. Educating the user to make them comfortable with this type of environment is important, and what they do with this device after they get it is critical.

I also want to say that Linux is absolutely mainstream across many devices. Set-top boxes, TiVo, all these electronics are loaded with embedded Linux. So it's just not visible. It's not merchandized or marketed. But Linux is extremely mainstream across many, many devices today.

Q: Can you describe a bit about the Nettop arena and how Moblin can take advantage of that market, and perhaps what some of the challenges are there?
Fisher: I think at a high level, the usage model is very, very similar to the Netbook. The actual design concept for the Netbook is being scaled because of the screen resolutions to take advantage of the larger screen resolution. The difference in my mind is more around input methods and the resolution of the screen. So there's a lot of innovation going on around the touch and gesture activities on the Nettop. I expect to see manufacturers in the future delivering the Moblin-type capabilities on these devices, as well as the media-phone-type devices and other things that really require an input method like a touch or gesture.

Q: You mentioned before there are already some Nettops running Moblin. Can you talk a little about that?
Fisher: Back in, I think, 2008, there was a big push to deliver Nettops by the OEMs, primarily in Taiwan. What we did was we inserted Moblin's capabilities into the releases, so that when these devices went out into the market, they had the Moblin technology, and capabilities were already available in them. We really dramatically reduced the power consumption with some techniques that we had driven into the Moblin project. We had boot-time capabilities that we had been working aggressively on. We made sure that they were incorporated into those releases. Fundamentally, those devices were released with Moblin technologies, and as I've said, now we've built a Moblin-compliant set of specifications, nearly completed, which really unifies and documents those things. And Moblin version 2 is where that really came together.

Q: What would Intel like to see from vendors and manufacturers?
Fisher: At Intel, our goal is to ensure that Netbooks, handheld devices, Nettops--all these are shipping with Atom. That's our pure goal. We want to be sure we ship Atom platforms. In order to do that, we have to have the best experience. Whether it's Windows or if it's Moblin, it doesn't matter to me.

Q: How does that play into Intel's working relationship with Microsoft in terms of what operating system ends up on these different platforms?
Fisher: At the highest level, we have a great relationship with Microsoft. They're delivering a wonderful operating environment. We're all looking forward to Windows 7. So there's no competition. The only competition we see is ensuring that as a hardware platform company, that we win in the platform space. So it's all about winning the Atom platform. We work very closely with Microsoft to ensure that Intel Architecture [IA] is optimized, that they take full advantage of the architecture and we take full advantage of the capability they deliver. And as the platform of choice, we're going to ensure the same thing when the Linux environment is chosen.

Q: Obviously one challenge of any operating system is getting people to write applications for it. Can you talk about the Atom Developer Program and your goals with it?
Fisher: On a high level, it's really about generating excitement around innovating on Atom. We think that by putting the development platform out there, we're going to get developers to innovate around these specific Atom platforms and then have a mechanism to monetize their efforts. And that's critical to creating demand in the ecosystems. So regardless of whether we're developing on Windows or Moblin, we want to see innovation. So we announced the ability to develop native Moblin applications. We have an SDK to develop natively to Moblin. Windows has always had a very strong presence in the development platform.

Q: I had read something about Moblin 2.0 supporting Android applications. Is that correct at this point?
Fisher: No, there are no plans for that.

Q: I had read some stories at one point that they're looking at supporting Android apps?
Fisher: There was a technology demonstration--open-source projects. There's a lot of innovation going on. But there's no product commitment. It's all community-type stuff. I think Ubuntu is doing some stuff.

Q: As far as the mobile device and notebooks as well, obviously Android is the other prominent player as far as open source on those different platforms. How does Intel see the arena between Moblin and Android? Is that an area of competition or more a peaceful co-existence between the two? Do you think the market can support both environments or is one going to have to be dominant over the other?
Fisher: It's a big market. Our focus at Intel, we're just maniacal on delivering the best platform for Atom, in this case the Atom version of our architecture. I'm focused on that. The Moblin effort is 100 percent focused from an Intel perspective. It's an open-source project. Our Intel effort is all around delivering the rich Internet experience, media consumption, social networking for these types of devices. A popular choice for Atom-based platforms is Moblin. So we are going to optimize the heck out of that. It's not about competition. For me, it's about selling Atoms. It's about us delivering the best experience on our Atom platform so that they sell.

Q: Moblin [2.1] is in a trial or beta stage right now. When is 2.1 scheduled to officially hit the market as a final release?
Fisher: Well, to hit the market is kind of a question for the OSVs and the OEMs. They are really the ones that productize and finalize it. From a community effort, we are expecting 2.1 capabilities in the October time frame for Netbooks so that then they can be productized. And then later on, late this year, first of next year, for the handheld-type devices, you'll see capabilities in the communities to productize those. We are really driving a timed market around our Moorestown platform for those types of things for 2.1.

Amid markedly lower sales and sluggish demand for its cell phones, Motorola saw a dramatic boost in earnings for the second quarter.

The company on Thursday reported net profit of $26 million, or 1 cent a share, for the quarter ended July 4, compared with $4 million for 2008's second quarter.

The jump in earnings surprised analysts, who had expected a loss of 4 cents a share, and Motorola itself, which had forecast a loss of 3 cents to 5 cents per share.

Quarterly sales fell to $5.5 billion, a drop of 32 percent from $8.1 billion a year ago. Analysts had been eyeing sales of $5.6 billion.

Ih the coming months, Motorola is hoping to make a dent in the smartphone market. "We have agreements in place with carriers and remain on track to bring our new smartphone devices to market for the holiday selling season," Sanjay Jha, co-CEO of Motorola and CEO of Mobile Devices, said in a statement. "We are also excited about our 2010 portfolio and are pleased with the customer feedback."

ZDNet's Larry Dignan noted the company's expectation of good results from tapping Google's Android software, which, Jha said, "will get us back in the game in smartphones."

"We will have two Android devices (for the holidays in the U.S.)," said Jha, adding that deals are already in place with two major carriers with more to follow. "We also have plans for more devices in the first quarter of 2010."

Facing a competitive cell phone market and the lack of exciting new products, the company has been struggling over the past couple of years. Motorola's popular Razr was unveiled in 2004. Since then, the company has failed to come up with a device hot enough to compete effectively with smartphones from companies such as iPhone maker Apple.

Motorola's 2009 second-quarter results

(Credit: Motorola)

During the second quarter, Motorola said it sold 14.8 million mobile phones, up slightly from 14.7 million in the first quarter but down from the 28 million units it sold a year earlier. Second-quarter sales for the mobile device division plummeted 45 percent to $1.8 billion.

To stay afloat, the company has been on a cost-cutting frenzy since last year, which contributed to the quarterly earnings gain.

In December, Motorola said it would trim compensation packages for all employees. Then in January, it announced staff layoffs of 4,000, around 6 percent of its entire workforce.

For the third quarter, Motorola is expecting anywhere from a loss of 1 cent per share to a gain of 1 cent per share. "We will further improve earnings in the second half of the year," said Jha on a conference call to announce the second-quarter results.

Intel has completed its purchase of Wind River, which builds software for smartphones and other devices.

The acquisition is costing Intel $11.50 a share in cash for a total price tag of around $884 million, Intel said Friday. Wind River is now a wholly owned subsidiary of Intel, reporting to the chipmaker's Software and Services Group.

By scooping up Wind River, Intel hopes to carve out a greater chunk of the mobile device market. Wind River designs operating systems and other software for cell phones, portable Internet devices, consumer electronics, and in-car "infotainment" systems. With such diverse customers as Sony, Verizon, Motorola, Boeing, and NASA, Wind River has its hooks in the automotive, aerospace, and telecommunications industries.

"The acquisition will deliver to Intel robust software capabilities in embedded systems and mobile devices, both important growth areas for the company," said Renee James, an Intel vice president and general manager of the company's Software and Services Group.

The purchase of Wind River also moves Intel further into software as an added source of revenue.

Though now owned by Intel, Wind River said it will continue to develop applications for its current customers. The company expects to pick up sales and new customers with access to Intel's technology, brand, and global sales force.

Intel first announced its decision to buy Wind River on June 4.

Intel and Nokia have more than a few holes in their respective collections of mobile technologies. How far will the collaboration announced Tuesday go to plug the holes and take them to the next technology plane?

Intel senior vice president Anand Chandrasekher

(Credit: Intel)

A platitude easily missed in the announcement may be the most revealing statement. Simply, that the two companies create the opportunity to take advantage of each other's expertise.

Nokia makes mobile phones. Intel, the world's largest chipmaker, can't get its chips into mobile phones. On the other hand, Intel makes the silicon that powers the world's PCs. Nokia doesn't have a clue about PCs.

The announcement won't necessarily inspire confidence with its lack of product particulars, but that's not what it's about. "Today is a relationship announcement," said Jeff Orr, senior analyst for mobile devices at ABI Research.

Intel and Nokia are simply agreeing at this stage to collaborate rather than be direct competitors, according to Orr.

Nokia was clear--in a cryptic sort of way--on one point, however: "Today's collaboration is not about smartphones but creating a new class of devices," Kai Oistamo, executive vice president for devices at Nokia, said in a phone interview Tuesday.

Beyond those future devices--presumably powered by Intel silicon--what does Intel get? Initially, the most concrete thing is 3G. "This is a gap for Intel, which has focused on Bluetooth, Wi-Fi, and WiMax," Orr said. "As a result, when future architectures like an Atom platform are developed for MIDs (mobile Internet devices), Netbooks, smartphones, that means vendors will have more flexibility for connectivity."

In short, Intel can build 3G into its chipsets and Intel can compete more effectively in the future with products like the iPhone and Palm Pre that include 3G as standard. Intel-based notebooks and Netbooks, until recently, were rarely offered with 3G as a standard option.

"We're not talking about specific products today but certainly we would not have taken a license (from Nokia) if we didn't have the intention to build a product," Anand Chandrasekher, Intel senior vice president and general manager at the Ultra Mobility Group, said in a phone interview Tuesday, referring to Intel's licensing of Nokia's HSPA/3G modem technology.

And it may be too soon for 4G technologies like WiMax. There are many countries (ABI Research's Orr counts about 100) where 3G is just emerging, so talking about WiMax (a 4G technology) is "very premature for most countries," he said.

Updated at 8:20 a.m. PDT: Added Intel-Nokia announcement and Intel discussion.

Intel and Nokia announced on Tuesday a wide-ranging deal covering chips, hardware, and software for mobile devices.

The companies said their new "long-term relationship" will focus on developing new chip architectures and software and a new class of Intel-based mobile computing devices. The move is part of a major shift for Intel, which is a giant in PC chips but not a player in cell phones.

Among other aspects, the agreement covers mobile applications and wireless Internet access "in a user-friendly pocketable form factor."

The Intel and Nokia effort includes collaboration in several open-source mobile Linux software projects. Intel will also acquire a Nokia HSPA/3G modem IP license for use in future products.

"We will explore new ideas in designs, materials and displays that will go far beyond devices and services on the market today," Nokia said in a statement.

For Intel, the deal adds momentum to its push into the small device/smartphone space. The Nokia announcement follows a pact announced with LG Electronics in February to collaborate on development of smartphones based on Intel's future "Moorestown" silicon and Linux Moblin software.

In March, Intel also announced a deal with Taiwan Semiconductor Manufacturing Company (TSMC) to cooperate in the manufacture of Atom processors.

Intel's need
The point of all of these announcements is to get Intel-architecture chips into cell phones, a giant worldwide market with well over a billion devices sold in 2008.

And the world's largest chipmaker needs to be a player in this market. Smartphones like Apple's iPhone, the Palm Pre, and T-Mobile's Google Android phone, the G1, are taking on many of the attributes of PCs and are increasingly adept at Web browsing, video streaming, and game playing--not unlike a personal computer.

Toshiba just began selling a smartphone that packs a 1GHz Qualcomm processor.

Texas Instruments and other chipmakers are also readying speedy processors for smartphones next year with two processing cores and enhanced video capabilities. And it was disclosed last week that an Nvidia chip will power Microsoft's Zune HD.

And what do those devices and technologies have in common? They're all powered by chips based on the ARM design.

Why ARM? ARM's approach to designing processors is the opposite of Intel's: power efficiency is paramount, performance secondary. Smartphone chips need to operate within a tiny power envelope, typically well under 0.5 watts and must last all day on one battery charge. Current Intel Atom chips--while relatively fast--draw too much power and are hardly suitable for smartphones.

The irony
Ironically, Intel manufactured an ARM-based chip series for many years called Xscale, which traces its heritage to a design called StrongARM. These chips were used in the Hewlett-Packard iPaq, a leading handheld for a number of years. But Intel sold this business to Marvell in 2006.

The chipmaker's strategy now is to shrink its global-standard x86 PC chip architecture to the point where it can run efficiently in smartphones. That's where Moorestown comes in. Intel claims Moorestown will be suited for high-end smartphones by 2010 and that "Medfield" silicon will make it into standard cell phones by 2011.

Neither Intel nor LG gave a date for availability of the LG device, but it is expected to appear soon after Moorestown is available. Intel is saying that Moorestown will be available in 2009 or 2010, though the second half of 2009 appears increasingly likely.

Under the agreement with TSMC, Intel will port its Atom processor technology to TSMC, which will serve solely as a manufacturer of Atom-related silicon--primarily chipsets.

ARM, the company that designs chips for the world's smartphones, said handset makers will deliver the first models using more than one processor core next year, as high-end mobile phones begin to mimic the hardware attributes of PCs.

Palm Pre

(Credit: Palm )

And that means top smartphone suppliers such as Apple and Palm, which both use ARM-based processors, will likely deliver models packing at least two cores inside the main processor--referred to as an application (or applications) processor.

This is a natural progression for high-end smartphones like the iPhone and Palm Pre as the software those phones run gets increasingly sophisticated.

"You'll definitely see handsets shipping with a dual-core A9 in 2010," James Bruce, wireless segment manager for ARM, said in a phone interview earlier this week, referring to the next-generation Cortex-A9 processor from ARM.

The Palm Pre uses a processor based on the current-generation Cortex A8. The iPhone also uses an Apple-branded chip that is based on an ARM design.

"The A8 is just a single core while the A9 will be dual-core, all the way up to quad-core to give smartphones an even bigger performance boost," Bruce said.

He said the move to dual-core phones should happen relatively quickly. "It's very aggressive. It's only going to be in a year's time that you're going to get these phones," he said.

And what about power consumption, a critical concern for smartphones looking to deliver all-day battery life? "What we've done on the A9 is actually make it more power efficient than the A8. The dual-core A9 will be coming out on 45-nanometer rather than the (current) 65-nanometer process," Bruce said. Generally, the smaller the geometries, the faster and more power-efficient the processor is.

Bruce continued. "With the dual-core running at maximum load there's probably going to be an increase of about 10 to 20 percent in power consumption but in general day to day use you're actually going to see better battery life."

Virtualization company VMware says that its new platform to let organizations virtualize smartphones and mobile devices will show up in handsets next year.

The mobile phone is now as important to businesses at desktop computers, and acts as a mobile computer in many cases, said Stephen Herrod, VMware chief technology officer, speaking at the Interop industry conference in Las Vegas on Tuesday.

Although Herrod said the project is still in its "early stages," researchers at VMware are working on a VMware Mobile Virtualization Platform, which will create a virtual machine for mobile devices, allowing users to move their phone to different handsets.

VMware is currently negotiating with handset manufacturers, and the technology should be included within new handsets at some point in 2010, said Fredrik Sjostedt, VMware EMEA director of product marketing. "The technology is there, and up and running, but it needs to be something that is on handsets when they're purchased," Sjostedt told ZDNet UK on Wednesday.

Virtualization of mobile phones means user data on phones will be a portable file that can be moved from handset to handset. This means users who lose or damage phones will be able to retrieve data more easily, and corporate IT departments will have fewer support headaches, said Sjostedt. Potentially, companies could also run more than one virtual smartphone on a single handset.

"Configuring e-mail, secure Web access and corporate applications on phones is a nightmare, particularly where users have gone out and bought themselves a new handset. This means you have an abstraction layer between the application and the phone's operating system, so you can deploy the same image on a Nokia or a Motorola handset, very easily," Sjostedt said. "Companies will be able to say--just bring your phone to us and we will create a corporate identity that can sit alongside your personal identity. Users can still have their own files and photos, without putting corporate security or data at risk."

Mobile virtualization will also mean handset manufacturers will be able to roll out new features and applications more quickly, Sjostedt added. At present, handset companies may spend 18 or 24 months testing new features on various platforms and phone models--but this could be eliminated using virtualized phones. VMware is currently in advanced negotiations with all the major handset manufacturers, Sjostedt said.

Mobile-phone virtualization is probably still an idea ahead of its time, according to Clive Longbottom, a research director with Quocirca. "At the moment, there's an element of 'what's the point?' but it's certainly interesting a year or two out," Longbottom said.

Longbottom believes the most valuable application of mobile virtualization will be in corporate settings, where it can be difficult to get full functionality from enterprise applications on mobile devices.

"I can see that a field engineer might find it useful to be working using a scheduling application, then have the ability to switch to his laptop image and see boiler plans and instructions when the need arises. Being able to divide mobile devices into different systems could potentially be very powerful," he said. "However, at the moment, I don't think devices like the iPhone are necessarily sophisticated or powerful enough to handle this kind of technology."

Sally Whittle of ZDNet UK reported from London.

Intel's Medfield is the chip that will drive the chipmaker's smartphone strategy in 2011 and beyond, according to an Intel executive speaking recently at an Intel investor meeting.

Slides (accessible on Intel's Web site) shown by Anand Chandrasekher, general manager of Intel's Ultra Mobility Group, at last week's Intel investor meeting map out the road Intel will take to the mainstream smartphone market. (The slides were highlighted on technology sites such as Engadget and UMPC Portal).

Intel is taking a graduated approach to the smartphone market: the ultimate target is the mainstream smartphone

(Credit: Intel)

Intel's Medfield was first disclosed in December.

Chandrasekher showed a slide that put Medfield in the mainstream smartphone market by 2011 by reducing the size and power requirements of the chip. "We take the power down again using 32-nanometer (technology), we, of course, take the performance up using 32 nanometer. But we also consolidated everything onto one chip and shrink the form factor (smartphone design) down again," Chandrasekher said at the investor meeting, which was streamed over the Web.

"We got to get the power down so we can get all-day battery life and get the (chip) package (size) down," he said. "It's not going to be (that in) one generation we fix everything," he added.

Chandrasekher said Intel's biggest advantage in the mobile Internet device and smartphone market is the most obvious: Intel's x86 architecture that runs the world's PCs.

He also offered some updates for Moorestown, the chip that will precede Medfield. "Last year I said we're going to do better then 10X on platform idle power (reduction over the current Menlow technology) on Moorestown. Today, I'm telling you, for the first time outside the walls of Intel, we're going to do 50X better on idle power," he said. The idle power will be 20 milliwatts at the "platform level"--which refers to the idle power (or standby mode) of the actual device, such as a smartphone, not the just chip, according to Chandrasekher.

Chandrasekher also clarified that Taiwan Semiconductor Manufacturing Company will be making the "Langwell" half of the Moorestown silicon. "That second chip (Langwell) we manufacture on TSMC because there's a lot of that IP (intellectual property) sitting on TSMC today," he said.

For comparison, he showed a current Google Android-based phone that had idle power of 20 milliwatts.