by Anton Shilov
02/20/2008 | 10:09 AM
The analyst who last week said that Advanced Micro Devices’ next-generation mobile platform may face problems with the launch retracted his comments, but said that notebook makers still remain cautious about the forthcoming launch due to track-record of AMD’s execution in the most recent quarters.
“[AMD] is not having technical issues with Puma, [but] thesis of cautious OEM behavior with respect to AMD’s road-map appears well founded,” Doug Friedman of American Technology Research wrote in another note to clients of the analyst firm, reports Barron’s.
Last week the observer of “AmTech” wrote that AMD code-named Puma mobile platform had technical problems based on “channel checks”, but the fix was in the works. Mr. Freedman did not specify what kind of problems AMD’s code-named mobile platform Puma may or may not have, whereas officials from AMD were quick to deny any issues with the forthcoming platform for laptops.
Nevertheless, the analyst still does not believe into successful and faultless launch of AMD’s Puma platform. He now claims that due to the lack of technical flaws the demand may exceed the supplies of AMD’s chips.
“[The laptop makers have contingency plans] in case they run into issues releasing new Puma-based products,” said Mr. Friedman about the scenario when AMD cannot meet demand towards its new products.
AMD’s Puma mobile platform will be based on AMD’s code-named Griffin processor, AMD M780 or AMD M780G mobile chipsets, which support PCI Express 2.0 bus, HyperTransport 3.0 bus, flash-cache for boosting hard drive performance and comes with or without built-in DirectX 10-compliant ATI Radeon graphics core that also features hardware-assisted decoding of high-definition content from Blu-ray and HD DVD discs. Besides, notebooks based on the Puma platform may be equipped with ATI Mobility Radeon HD 3000-series discrete graphics processors as well as 802.11 a/b/g/n wireless network controller as well as 3G option.
It is necessary to note that AMD’s code-named Griffin microprocessor is immune to the notorious TLB-bug with AMD Phenom chips due to the fact that Griffin central processing unit is based on the K8/K9 micro-architecture with many tweaks made to HyperTransport controller, DDR2 memory controller as well as with aggressive tuning to save as much energy as possible. The chip will be made using proven 65nm process technology.