AMD’s Griffin, Puma Pounce on Mobile Market

AMD executives began spreading the word Friday about the company’s code name for its new next-generation mobile microprocessor, named “Griffin,” as well as its new platform for notebook computers, known as “Puma.”

Expected to debut in laptops in the latter half of 2008, the chip and platform will go into production later this year.

While Intel has launched a series of chips since 2003 designed specifically for notebook computers, Griffin is AMD’s first chip to sport an architecture specially designed for mobile computing. With Puma, AMD has a platform positioned to directly compete against Intel’s Centrino.

“With this introduction, AMD is stepping up its mobility story,” said Roger L. Kay, president of Endpoint Technologies Associates. “With increased performance and power efficiency, Puma represents the company’s first explicitly mobile platform. It’s safe to assume that this offering is just the first of what will turn out to be a stream of evolving products based on the company’s new open platform mobile technology.”

Going Somewhere

The Griffin processor’s new architecture bumps up AMD’s current offerings with a dual-core chip. The chips will be manufactured using the company’s 65-nanometer process.

The two cores and the integrated memory controller will be built on separate power planes. This will enable, for instance, each core to go into deep sleep states independently as the remaining core goes about its work, ultimately reducing power consumption and increasing battery life. The newly redesigned integrated memory controller will also increase dynamic RAM (DRAM) efficiency.

These improvements may be good news for AMD, but with Intel in the market for the past four years and continually improving its line of mobile processors, Griffin has left some analysts flat.

“The power use and the processing cores — that’s going to help. And they are also incorporating the deeper sleep mode,” Jim McGregor, principal analyst at InStat, told TechNewsWorld. “But then when you look at it, Intel already has that level of power management granularity and even more so.

“[Intel is] now adding an even deeper sleep, and they’ve even taken things farther so that they can overclock one core if the other core isn’t being used, and AMD still does not have that capability,” he added.

Platform Dive

With the Puma platform, based on the Griffin chip, and the RS780 chip set featuring PCI Express Generation 2 and HyperTransport 3.0, AMD benefits from its purchase of ATI with a chip set the company said will deliver a rich visual experience and increased performance.

The platform features motherboard DirectX 10 graphics processing, energy efficient high-definition multimedia support with the Unified Video Decoder, integrated multi-monitor support with DVI (digital video interface), HDMI (high definition multimedia interface) and DisplayPort, native Southbridge support for NAND flash with HyperFlash and PowerXpress for dynamic switching between integrated and discrete graphics for extended battery life.

“Obviously Intel has been upping the game with their new processors. They are adding features, but they are not adding anything Intel does not already have and more,” McGregor pointed out.

“Other than some new features on the processor side,” McGregor said, “the innovation is kind of minimal at this point. It’s more of a small evolutionary change rather than a significant jump in performance.”

Leave a Comment

Please sign in to post or reply to a comment. New users create a free account.

Cryptocurrency is a viable hedge against inflation: Agree or Disagree?
Loading ... Loading ...

LinuxInsider Channels