It explains such things as the Genetic computing, floating buffer and other features found in the Articia chips.
As a side note, there is mentioned, page 24, that the Articia Sa/P can handle a 166Mhz fsb when being produced on a 25nm process. As one can allso buy masks from MAI designed for a 18Ám process [page 10], I don't think it's to unlikely that those could use a faster fsb(200Mhz) and memory controller(400Mhz), and perhaps a faster PCI bus(PCI bus 2 running @ 66Mhz, on the Articia P that is).
The faster the FSB and Memory Controller is, the better: this is the real bottleneck in daily power-computing. Doubling the speed of the current procesors has not such a big benefit as doubling the speed of the mainboard and memroy would have.
Re: Read a technical summary of the Articia chipsets Posted on 20-Nov-2003 20:24:56
Oh I dunno. I suspect the PC architecture not being able to walk and chew gum (e.g. follow PCI interrupts *and* refresh the display) at the same time has a lot to do with it.
Make the computer properly multitasking and you don't need to scrabble for the tiniest speed increase elsewhere.
muttermutter power computing - exactly that - vast amounts of power required to get a halfway decent result when better-designed machines can do the same thing for half the cost (or even less) muttermutter
Re: Read a technical summary of the Articia chipsets Posted on 21-Nov-2003 9:44:38
18nm process??? I though the 90nm process is the smallest process that will be used in the near future. Even Intel and AMD and all the other fabs have still problems with this process! A 18nm process will be used in 2025
I think you mean the 0.18Ám process. Is this right?