Payoff time for Intel: The Centrino 2 platform goes live

By Scott M. Fulton, III | Published July 15, 2008, 12:06 PM

What Intel has been calling its "tick-tock" cadence almost lost its "tock." But what was expected to be a June release ended up being July, giving partners plenty of time to build up Centrino 2 inventories in time for back-to-school.

After a delay of what ended up being only a handful of weeks, Intel's Centrino 2 mobile technology platform is now debuting worldwide. It's based on the company's 45 nm "Penryn" generation processors unveiled earlier this year, including the all-new Core 2 Duo T9400; and the new Mobile GM45 Express chipset, which boosts the memory bus speed up past the 1 GHz mark.

So what does this mean? Next year, when Intel begins premiering its "Nehalem" generation processors, its fundamental technology will change and there will no longer be a separate memory bus. For now, a boost to a 1066 MHz memory bus means that memory transfers will take place under a 60% faster clock than the current generation.

Arguably, that affects more this year than it would had the technology been introduced last year. In recent months, we've seen a shift in manufacturers' marketing of notebook computers toward portable multimedia (especially with the new 1080-line displays at the high end), which means that gaming is being nudged aside, becoming a more marginal segment of the market. Video is taking its place, and video codec performance becomes key. That's the application that memory bus speed boost will benefit most.

And as we see more manufacturers choosing to implement GPUs (as an alternative to Intel's built-in GMA) that implement shared memory -- where part or all of graphics memory is apportioned from the main memory pool -- that main bus speed boost will help compensate as well. Windows Vista users especially have complained that graphics processors allocating memory from the main pool contributes to system slowdowns; those slowdowns won't be as bad with 1066 MHz.

Intel's Mobile GM45 Express chipset

We will probably see more laptops in the midrange market segment that boast multimedia performance over 3D rendering performance -- in fact, there's a growing segment that manufacturers are tapping into now, for whom 3D isn't even an issue. The new Intel chipset plays right into their hands, with its new 4500 MDH Graphics Media Accelerator. Yes, it's a faster processing core at 533 MHz, but it will still play to that market segment that uses 3D only occasionally.

And with competitors NVidia and ATI both targeting the consumer video/multimedia segment more and more also, it may become more difficult for them over time for systems with discrete graphics chips added on, to justify their value proposition. Why pay more when Intel's integrated graphics seems to process video fast enough?

Notice that Intel's major new processor announcement this week is the T9400 -- a midrange processor that doesn't follow the trickle-down theory of technology adoption, where the best products start at the high end and work their way into the midrange after one year. With its "Conroe" desktop processor launches in 2006, Intel learned that it can benefit by tailoring the CPU that sits in the sweet spot for price/performance, to the sensible buyer who looks for that sweet spot. Specifically, it priced the Core 2 Duo E6600 at a level that literally destroyed AMD's value proposition in the midrange.

This week, Intel is gearing up for a repeat performance, targeting its new dual-core 2.53 GHz T9400 at $316 (at purchases of 1,000 units and up), while its 2.8 GHz T9600 is priced at $530 in quantity. There you see Intel hammering that sweet spot, and that dent will surely be reflected in new Centrino 2 notebook prices.

Centrino 2 is expected to be the platform where Mobile WiMAX makes its entry onto the world stage. But in this country, at least, Mobile WiMAX isn't ready for Centrino 2, except for tests in a meager few US cities. In fact, at this rate, the Nehalem-based mobile platform will probably already be at the high point of its market lifecycle before services such as Sprint's Xohm are available to a wide enough segment of the population to make it viable for business travelers.

Let's face it, the real purpose of a mobile technology platform is to make it easier for manufacturers to build systems. When AMD first combated Intel's original Centrino platform with its message of "choice" and "options," it learned to its dismay that manufacturers don't really want options. They prefer the stability of knowing that what they assemble is guaranteed to work. That's why AMD is emerging into platforms from the backdoor now, starting with Puma.

With the Mobile GM45 Express chipset selling to manufacturers at quantity for $43, and with Centrino already having been developed by Intel into a trusted brand, Intel has a very real possibility of clinching the mobile platform market for itself. That would drive AMD back to being the "alternative choice," which is a much more comfortable position for it to be in...from Intel's perspective.