What do you want in a mobile computer?
How much performance do you want to give up for longer battery life? Would you buy a clunky mobile computer that can run anything you throw at it? If you're the envy of the digerati when you walk down the street with your new phone, but you can't use it to make reservations at Nobu, are you still cool?
Chipmakers are struggling with these questions as well as how to adjust their recipes for the future of mobile computing. It's not so much the about chips themselves, but how you use them.
The current model for smartphone chip design used by companies like Texas Instruments, Samsung, and Freescale Semiconductor is to minimize power consumption by relying on an ARM applications processor for many tasks, but spreading much of the work around smaller components called accelerators that are activated only when needed.
To get an iPhone-like combination of battery life and sleek design, ARM's partners use several specialized chips.
(Credit: CNET Networks)Intel acknowledges the special role that hardware accelerators play in this market, because a modern general-purpose processor like the Core 2 Duo simply won't fit into a handheld device. But it argues that the role of accelerators should be somewhat limited to maximize software compatibility, and that the processor, such as its low-power Silverthorne and Moorestown designs, should be able to take on a great deal of work.
So which approach is best-suited for the next generation of mobile computers? Modern smartphones offer snappier performance and can run all day, but the increasingly sophisticated software that people want on these devices can overwhelm the system and cause compatibility issues. Running a wide variety of software won't be a problem for devices based on Silverthorne and Moorestown, but Intel still has a ways to go to allow its customers to build devices that are design showpieces like the iPhone.
One of Intel's most important goals over the next five years is to establish itself as a silicon supplier to future manufacturers of ultramobile computers, whether those are smartphones, Mobile Internet Devices, UMPCs, or some other design that emerges from that soup. It's already tried this once, attempting to break into the mobile phone market with an ARM-powered chip that was a model of integration, but a flop with customers.
ARM and its partners enjoy almost complete control over the current smartphone market and are eyeing the evolution of the platform just as closely. This is the next growth engine for personal computers, something made extremely clear last year with the steps taken by outsiders such as Apple and Google to gain a foothold.
Eye on power consumption
The mobile phone chip suppliers evolved their strategy by focusing first and foremost on power consumption, a luxury they were afforded by the relatively simple software run by most mobile phones up until fairly recently.
The basic idea is that most of the time, your phone or mobile computer isn't doing a whole lot. But every now and then, you need a performance boost to fire up a video clip, download something from the Internet, or take a picture. In the ARM community, chip companies deliver this performance with a complicated concoction of multiple processing cores that only handle one specialized task, such as video encoding.
"What happens is that each task runs on the specific engine that is much better optimized from a power generation standpoint," said Avner Goren, worldwide director of cellular systems market for TI. "Offloading video from a central CPU to a video and imaging accelerator decreases the power generation."
Intel is used to integrating as many of those kinds of functions as possible into its PC chips and might have been expected to bring that same approach to this new battleground. But as it prepares to roll out Silverthorne, and later Moorestown, power consumption requirements are forcing Intel to acknowledge that it can't do everything with a single chip.
"We don't expect everything to be done by the general-purpose processor," said Ticky Thakkar, director of Intel's ultramobile group platform architecture and an Intel fellow. "(But) you need it for compatibility reasons; you need the same software that runs on a PC to run on a handheld device."
Intel hopes device makers will build sleek and powerful designs with its Moorestown chip
(Credit: Tom Krazit/CNET News.com)As with many decisions in chipmaking, the question is one of engineering trade-offs. ARM, TI, and the other smartphone chip companies are trying to step up to build more powerful computing devices, while Intel is trying to cram a PC into a handheld. Their respective starting positions dictate the philosophies to a great degree.
Why should you care? TI argues that its approach results in better battery life and can actually improve the responsiveness of the hardware, since the various accelerators band together to spread out the load.
"When you are using an architecture of multiple cores and multiple engines, and you offload the multimedia tasks to the coprocessors, the ARM (core) is more responsive because it's not loaded," TI's Goren said. This approach also has power benefits that allow device builders to push the envelope of slim and stylish design.
Thakkar concedes that point, but notes that a more powerful CPU can handle popular software-based codecs like Flash or Javascript that don't need an accelerator. Flash is hard to find on a mobile phone. The alternative is Adobe Systems' Flash Lite technology, which isn't bad but isn't the same as the PC version of Flash.
This is the heart of Intel's pitch to mobile device makers like Nokia and Apple: we'll be able to support anything you currently enjoy on your PC, on your mobile phone. Not necessarily at the snappy performance levels you're used to on a Core 2 Duo laptop, but Intel wants to deliver a PC-like experience on a mobile phone.
A smartphone is no PC
As advanced as smartphones have become, they're not overtaking the PC on any benchmarking applications anytime soon. And as Intel's manufacturing operation continues to crank out smaller transistors, and integrates more and more of them onto a processor, Thakkar thinks it's going to be hard for other players to match Intel's combination of software compatibility and performance.
But then again, is that what you want from a mobile phone, a little PC? "A laptop is not a cell phone. When I boot my laptop, I've trained myself to go get coffee. When I switch on a cell phone, I expect that I can make a call in seconds," Goren said. Running PC software might be nice in theory, but TI and the mobile phone companies think that software written for the PC is too bloated for the phone environment.
This isn't a one-size-fits-all market. Consider just how many types of mobile phones there are, from iPhones to Voyagers to BlackBerrys to Hiptops. Those phones all require a different balance between performance, power consumption, battery life, and software requirements. But for the most part, the laptop form factor isn't all that much different between a Pavilion, an Inspiron, or a MacBook.
Phone makers like Nokia, Motorola, and Apple will decide how this battle plays out. For smartphones to really evolve into mobile computers, they'll need to ensure that software written for one operating system will run across many devices--which isn't always the case now and is a large reason why Google is getting involved. But for Intel to find the elusive source of growth outside the PC market it so desperately needs, it's going to have to deliver on the audacious goals it has set for power consumption.
Silverthorne, due out in the second quarter, isn't quite the answer. Moorestown, which is expected to arrive around 2010 and consume 10 times less power than Silverthorne, could be. If Intel can pull that off, it would put Moorestown down on the milliwatt range currently occupied by ARM's chips, with performance that should rival anything from TI, Samsung, and the like.
"We've made these jumps where we've taken order-of-magnitude leaps," Thakkar said. For Intel to find the next big source of growth that has eluded its grasp this entire decade, it will need to do that again.
Intel filled in some of the missing details on its Silverthorne mobile processor Tuesday, helping explain how it managed to get the power consumption of this chip down under a couple of watts.
Gianfranco Gerosa of Intel presented the company's paper on Silverthorne, its low-power mobile processor destined for the next generation of mobile Internet devices later this quarter, during the International Solid State Circuits Conference in San Francisco. Intel Chief Technology Officer Justin Rattner had already discussed Silverthorne in some detail last week, but the wonky details were laid bare for a roomful of people who are way, way smarter than me.
A few Silverthorne tidbits, however, could be deciphered by those of us who wasted their education dollars on a business degree. This chip is tiny, measuring just 3.1 millimeters by 7.8 millimeters for a die size of 24.2 millimeters squared. By comparison, the dual-core version of Intel's newest Penryn chips for PCs has a die size of 107 millimeters squared. That means Intel can make roughly four times as many Silverthorne chips on a single silicon wafer as compared with the dual-core Penryns. Let's see how much the company decides to charge for it.
The chip will be able to reach 2.5GHz, although Intel is quoting a 2GHz clock speed for the 2-watt thermal design power, or the maximum power consumption that system designers have to take into account when building their devices. It uses a 16-stage pipeline, compared to the 14-stage pipeline used by the Core 2 Duo chips.
Think of a pipeline stage as part of an assembly line: the more stages in the process, the faster it has to run to build something in the same amount of time as a line with fewer steps. Intel's Pentium 4 processor topped out at 31 stages, which allowed the company to crank it up over 3GHz to satisfy the marketing department's proclamation that the only thing us mortal PC buyers understood was clock speed.
Unfortunately for Intel, that wasn't a very sound design. A chip running at that kind of speed runs way too hot, especially as current leakage problems became more pronounced, so Intel designed the Pentium M microprocessor with fewer pipeline stages. It did more work per stage, which allowed it to run slower and cooler. Eventually, those design principles were incorporated into the Core lineup of processors, and Intel got its mojo back.
So it's a bit surprising that its most serious low-power effort to date would have two additional pipeline stages, but Intel got around that problem by switching to an in-order pipeline, and by adding hyperthreading. That combination produced the most efficient performance-per-watt ratings in Intel's internal testing, Gerosa said.
Silverthorne also makes use of several low-power states in which the chip shuts down certain elements of the processor when they aren't required by the software. Intel estimates that Silverthorne will spend 90 percent of its time in deepest sleep state, which it calls C6. Virtually everything gets turned off in C6, and it takes 100 microseconds to wake the chip back up when new processing orders come in, Gerosa said.
As a result, Intel is quoting average power numbers for Silverthorne "in the order of a few hundred (milliwatts)," which sounds like quite the accomplishment. That will be nice for battery life, but it doesn't really matter when it comes to building a sleek device. Anyone who wants to use Silverthorne will have to design a device that can handle the full 2 watts of power that Silverthorne will consume running flat out.
After all, the whole point of Intel's pitch to put x86 chips in mobile devices is that those devices would be able to run Windows and any piece of PC software. While Intel is increasingly pitching Linux for its mobile devices--and keeping a close eye on that other mobile operating system--there are certain tasks that are going to require all the processing power Silverthorne can deliver.
The MIDs that Intel and its partners have shown off using Silverthorne don't look all that different from the older MIDs that haven't sold very well to date. They're still a little too bulky to compete with slicker smart phones from the likes of Nokia, Samsung, and Apple, which are powered by chips designed by ARM for mobile phones.
True competition from Intel in this area probably won't arrive until the Moorestown chip is ready in a couple of years, but Silverthorne is a milestone on that path. We'll start to see if people are interested in MIDs based on the chip by the middle of this year, when we'll also get a true sense of its performance.
Intel is expected to shed light on its processor of the future this week as it plugs along with another design that was once supposed to be its processor of the future.
The chip industry's finest minds will be descending on San Francisco this week for the International Solid State Circuits Conference, and Intel plans to present 14 papers highlighting some of its recent work, said Justin Rattner, Intel's chief technology officer and head of Intel Labs. Chief among them will be its low-power Silverthorne processor, Intel's latest plan to infiltrate the world of handheld devices.
"This is the smallest (x86 instruction set) processor we've built in the last 15 to 17 years," Rattner said, speaking of Silverthorne in a briefing for reporters prior to the conference. Silverthorne is expected to arrive in the second quarter in so-called mobile Internet devices, but it's really a stepping stone for Intel toward a lower-power future.
The ultimate goal for Intel is to design a chip that will fit into the next generation of mobile devices and smartphones, carving out a niche for PC software and programming techniques in a world dominated by chip designer ARM and its partners, such as Texas Instruments and Samsung. With a power consumption range between half a watt and two watts, Silverthorne doesn't quite fit into the appropriate thermal profile occupied by those chips. But it can run any piece of software written for PCs at the performance level achieved by one of Intel's 5-year-old Pentium M processors.
Silverthorne is a departure from the rest of Intel's designs in that it's an in-order processor, as opposed to the out-of-order design used in just about every other chip the company makes. That means pretty much what it sounds like; in-order chips have to process tasks in a defined order, but out-of-order chips can process tasks separately and reassemble a finished product later. Out-of-order means better performance, in-order generally means lower power consumption.
The chip also incorporates a new low-power state, allowing it to essentially shut down in between processing tasks and limit power consumption. And just in case the software industry starts producing more multithreaded applications, Intel added an old friend, hyperthreading, to Silverthorne. Hyperthreading allows Silverthorne to present itself as a dual-core chip to software, even though it only has one physical processing core.
While Silverthorne is considered a huge part of Intel's future, Tukwilla, the other major design being presented at ISSCC, is a remmant of Intel's past. Tukwilla will be the next version of the Itanium processor, which Intel once thought could take over the server industry but has been relegated instead to a high-end niche.
Still, that's a powerful niche. Tukwilla will be the first quad-core Itanium processor, and it will use a whopping 30MBs of cache memory. This is going to be a high-performance (read: expensive) chip for high-end servers, and it will also be one of Intel's first chips that borrows design techniques employed by Intel's rival, Advanced Micro Devices, to great effect.
Tukwilla will use point-to-point connection technology that Intel calls Quickpath to directly link processing cores together. AMD introduced this way back in 2003 with its Opteron processor, and it delivers a significant boost in performance as the chip industry moves into the multicore era. Intel's current designs required signals traveling between two different processor cores to leave, then re-enter, the chip. Quickpath lets those signals stay on the chip, allowing them to travel a shorter distance at higher speeds. That's good.
Tukwilla will also use an integrated memory controller, which is a similar concept. Integrated memory controllers, as the name implies, are integrated directly onto the chip. This allows this vital link between the processor and the main memory to run at the speed of the chip, rather than the slower speeds necessitated by the front-side bus design used by Intel's chips.
Intel eventually plans to bring Quickpath and integrated memory controllers down to its Xeon line of server processors, which go into the vast majority of the world's servers. That will arrive with the Nehalem generation of chips, due out later in 2008.
Intel's researchers also plan to present papers in other areas, such as wireless technologies and memory research. The company has been working on radios that can employ multiple protocols, from Wi-Fi and WiMax to even cellular standards, and it plans to highlight some research in this area. The company is also investigating phase-change memory, one possible way to prepare for the end of Moore's Law by discovering ways to represent the fundamental zeros and ones of computing with something other than a transistor.
These presentations are not for the faint of heart. An electrical engineering degree would be very helpful in deciphering the papers, although not exactly required. We'll be on hand during the week to highlight some of the more interesting presentations, especially Silverthorne, which is a key part of Intel's future plans.
After holding off on the release of a faster iPhone because of concerns about battery life, is Apple really prepared to take a step backward with Intel's Silverthorne chip?
AppleInsider reported Friday that Apple has decided to use Intel's upcoming low-power Silverthorne chip in "not one but multiple products currently situated on its 2008 calendar year product roadmap." Silverthorne is Intel's latest push to capture the handheld/mobile phone market as part of a product concept called the Mobile Internet Device.
It's probably a little too soon for an iPhone based on Intel's chips, but maybe not another kind of handheld.
(Credit: CNET Networks)The report goes on to say that the most likely candidates for Silverthorne are a 3G iPhone and the Newton-like tablet computer that AppleInsider reported on earlier in the year. Based on what we know about Silverthorne, I think the subtablet rumor might make sense, but a Silverthorne iPhone is unlikely.
Intel plans to release more details about Silverthorne at the Intenational Solid State Circuits Conference in February, but we already know from the advance program, and from what Intel said about the chip last year, that we can expect Silverthorne to behave like a much smaller 2004-era Pentium M processor that consumes just a watt or two of power, compared with the 35 watts consumed by Intel's Core 2 Duo notebook processors of today.
But that's still not enough for a phone. According to several iPhone teardowns, Apple is likely using the Samsung S3C6400, or some special equivalent built just for them, in the iPhone. That chip is based on the ARM1176 core, which at 620MHz consumes just 279 milliwatts. That's running all-out, whereas most of the time you're actually going to be drawing much less power than that. Silverthorne, by contrast, will consume 500 milliwatts of power at minimum, and probably only when it's doing nothing in idle mode.
Those numbers just aren't going to work in a phone, especially an Apple phone, if the company really is so concerned about power consumption that it has held off on releasing a 3G iPhone until the power consumption of that modem improves. Those numbers could work, however, in something more along the lines of a powerful handheld such as the rumored "Return of the Newton" that was discussed earlier this year. Based on the concept designs shown by Intel last year for Silverthorne-era devices, however, this would be something much larger than a phone, more along the lines of a UMPC or a handheld gaming device like Sony's PSP.
This is what Intel has in mind for its 2009 Moorestown chip, not its 2008 Silverthorne chip.
(Credit: Tom Krazit/CNET News.com)Given the close relationship between Apple CEO Steve Jobs and Intel CEO Paul Otellini, as well as Otellini's commitment to low-power designs, I would not at all be surprised to see Apple and Intel hook up on a future mobile phone or sleek mobile computer. But I wouldn't expect to see it until at least 2009, when Intel releases a chip called Moorestown that is expected to reach the milliwatt operating power of current ARM designs. Some of the concept devices that Intel showed off as Moorestown-era projects looked an awful lot like the iPhone.
Apple had to design the OS X operating system inside the iPhone and the iPod Touch around the ARM instruction set, because there's really no other realistic option right now for smart phones. But it might be looking at the development resources needed to port all of its software (iLife, iWork, GarageBand, etc.) over to ARM, and balking at the amount of time and energy that would require. If Intel can deliver an x86 chip with similar/better power and performance characteristics to the chips built by ARM's partners, it could be a very attractive product, and that's the meat of Intel's sales pitch to the phone makers.
Of course, ARM's partners will be out with dual-core chips by then that could tilt the performance equation solidly in its favor, while staying within the same power envelope. We've got a long time to muse about that.
Two weeks after Intel signaled its future low-power intentions, ARM has unveiled its latest mobile chip design for smart phones and consumer devices that will arrive around 2010.
ARM CEO Warren East, left, and marketing executive John Goodacre discuss the launch of the Cortex A9 at the ARM Developers Conference.
(Credit: Tom Krazit/CNET News.com)The Cortex A9 is an extension of the Cortex family of applications processor cores that ARM unveiled two years ago with the Cortex A8. It combines the multiprocessor support of older ARM cores with the Cortex design, ARM's highest-performance implementation to date. Several ARM partners, such as Texas Instruments, Samsung, STMicroelectronics, Nvidia and NEC Electronics also announced plans to use the Cortex A9 in future chips for smart phones and consumer electronics devices.
ARM, based in Cambridge, England, doesn't actually make chips. It designs processor cores that companies like TI and Samsung use in smart phones made by Nokia and Apple, respectively. There's an ARM core in more than 90 percent of the mobile phones in the world, and in many cases there are several ARM designs inside your phone.
"The ARM world is growing a lot faster than the economy as a whole, and the semiconductor industry as a whole," said Warren East, ARM's CEO, in a press conference Wednesday at the ARM Developers Conference. Smart phones aren't nearly as prevalent as PCs, but they are growing much stronger than their larger cousins, and ARM is the predominant chip architecture used in those phones.
As a result, Intel wants a piece of this market as it evolves. At some point down the road, either smart phones are going to become more sophisticated, or minitablet PCs are going to become sleeker and offer better battery life. Both Intel and ARM are positioning themselves to be inside future mobile computers, and each brings different strengths to the table.
ARM's John Goodacre, program manager for multiprocessing, said mobile chip makers will be able to implement up to four processing cores with the Cortex A9. He doesn't anticipate that smart phones even around the end of the decade will need that much performance, considering that PC customers today are having trouble justifying four cores. But embedded devices like in-car processors and networking gear, where ARM customers also build chips, will be at those levels in 2010 or thereabouts.
In the most powerful configuration, Cortex A9 chips should be able to deliver up to 8,000 DMIPS (dhrystone million instructions per second) of performance with power consumption of around 250 milliwatts. DMIPS is an older measure of integer performance that's mainly used these days for embedded chips that don't run nearly the amount of code that PC and server chips have to handle, so it's difficult to judge exactly how much performance that is compared with a modern-day PC processor from Intel or AMD.
But 250 milliwatts of power consumption is far below what the most power-sensitive PC chips are capable of delivering these days, and devices with ARM chips are already at that level today. The ARM11 core, which is the basis for the Samsung applications processor used in the iPhone, has around that level of power consumption.
Intel hopes to get well below a watt with its Silverthorne processor, due out next year. And in 2010 it plans to ship a processor called Moorestown that will probably compete directly against the Cortex A9 for design wins in future smart phones or MIDs, Intel's vision of the future of mobile computing.
In fact, Digitimes reported Wednesday that Apple is considering replacing the ARM-based Samsung chip currently found in the iPhone with Moorestown once the product is ready. Intel has released little information about Moorestown, other than to say it wants to dramatically reduce the power consumption of its products by the time that arrives around 2010.
UPDATE: An ARM representative pointed out that the Cortex A8 was actually the first Cortex applications processor for products like smart phones. The first Cortex processor was the Cortex M3, designed for networking gear and other embedded devices.
Apparently the Newton doesn't fall very far from the Apple tree.
AppleInsider is reporting that Apple has a project underway to develop a minitablet computer based on the OS X operating system it has developed for the iPhone and the iPod Touch. The site is calling it "the return of the Newton," referring to the much-maligned but cult-favorite PDA that Apple sold in the mid-1990s.
Apple's design ethos has advanced a bit since the original Newton, but could the company have a successor in the works?
(Credit: Apple)The report describes a "slate" style device that's about 1.5 times bigger than an iPhone, with a high resolution display and the same touch-screen interface that's found on the iPhone. AppleInsider thinks we might catch a glimpse of this device at Macworld in January.
If true, this would appear to be Apple's take on Intel's MID (mobile Internet device) concept, rather than a PDA like the Newton. There's simply no market these days for the traditional PDA, as even basic mobile phones can do everything a PDA can do, just with more style. But there's not a huge market for UMPCs or MIDs at this stage of the game, either.
Apple's OS X and multitouch interface are definitely unique takes on the UMPC/MID concept. Many of the devices demonstrated by Intel and its partners last week at IDF required a stylus for navigation, reminiscent of PDAs. The wide-screen slate design reported by AppleInsider would allow for some interesting applications. This would also have a larger screen than either the iPhone and iPod Touch, and the size of the screen could make it easier to use the touch-screen keyboard in landscape mode, something that's on the wish list of many iPhone users.
Momentum is building around mobile devices designed for everyday folks, but I'd still be surprised if Apple were to launch its third multitouch device in 12 months sometime next year. Is there enough interest in handheld Apple gadgets right now to justify the iPhone, the iPod Touch and Newton 2.0?
The timing reported by AppleInsider--release in the first half of 2008--does coincide with the launch expectations for Silverthorne, a low-power Intel processor designed for just this type of device. Intel executives were tight-lipped last week about their chances of working with Apple on Silverthorne-based devices, probably not wanting to get sent to the principal's office for talking out of turn about Apple.
SAN FRANCISCO--Intel's a big company, with lots of money and smart people. It will need both to take over two separate industries.
The company's official search for the next big thing is settling quite definitively on mobile computers. But this is actually two big things: not only does Intel want to create an entire new category of handheld computers called Mobile Internet Devices, it wants to set up a whole new network to service those devices.
Intel executives Dadi Perlmutter and Anand Chandrasekher laid out the company's mobile strategy Wednesday here at the Intel Developer Forum during a pair of morning speeches. For the past several years, Intel's mobile strategy has centered around notebooks, but it's eyeing more than just PCs these days.
Intel firmed up plans to ship Silverthorne, a new processor with lots of integrated features, next year. Chandrasekher showed off prototype devices built on Silverthorne that look an awful lot like the UMPCs that didn't exactly fly off the shelves: larger than a smart phone, dependent on a stylus, and many with a fixed keyboard. He also discussed a newer low-power concept called Moorestown that will consume 10 times less power than Menlow, the platform that will house Silverthorne. Silverthorne itself consumes 10 times less power than the original Banias Pentium M chip, Chandrasekher said.
Intel is lining its Silverthorne chips up against the mobile industry's ARM-based processors in anticipation of the next silicon battleground. Having tried and failed to get its chips inside mobile phones, Intel's now trying to drum up demand for MIDs as a smart phone alternative to getting the Internet in your pocket, as Chandrasekher put it. He must have bigger pockets than I do, because none of the MIDs shown onstage at IDF would fit comfortably in my pocket.
The company will get a little closer to that goal with Moorestown. Chandrasekher didn't say much about that chip, but showed off a concept device that bore a striking resemblance to an iPhone that had been stretched lengthwise. Just a coincidence, I'm sure.
But whatever Intel's partners build with its mobile chips, they'll have to somehow connect to the Internet. That's where WiMax comes in. Intel has been talking up WiMax for several years as an alternative to cellular networks and to the expense of wiring the world with high-speed cable. It's finally getting ready to start testing the waters with its mainstream products.
Next year, Intel will refresh its notebook technology with a product called Montevina. That will come with a Penryn processor, a new chipset, and an integrated Wi-Fi/WiMax radio that will come as an option as part of the Centrino brand.
It will be interesting to see how many people opt for the WiMax radio, or even whether they know they have WiMax service available. Later today I'll get a chance to check out the devices themselves, and ask more questions about Intel's strategy for MIDs and WiMax. Stay tuned.
- prev
- 1
- next





