AMD's R600 Die Revealed
Winga
MrSouth Africa Icrontian
Details are beginning to surface regarding AMD's first stand alone GPU marchitecture.
The chip location interestingly, is rotated at a 60 degree angle. Notably, this is not the first case of GPU manufacturers rotating the chip on the packaging, however this is done for two good reasons:
Firstly, AMD needed more room for resistors, which are clearly evident around the big die.
Secondly and more important, to shorten the traces to video memory as much as possible, in order to reduce the EM noise produced by the PCB.
Industry news sets the R600 launch for January 20th, 2007
Source: The Inquirer
The chip location interestingly, is rotated at a 60 degree angle. Notably, this is not the first case of GPU manufacturers rotating the chip on the packaging, however this is done for two good reasons:
Firstly, AMD needed more room for resistors, which are clearly evident around the big die.
Secondly and more important, to shorten the traces to video memory as much as possible, in order to reduce the EM noise produced by the PCB.
[img]http://www.short-media.com/images/newsimages/2006/November/R600 CPU.jpg[/img]
When compared to the R580 chip, we note that the R600 die is larger by 15 to 20 percent. Given the fact that the R600 is being manufactured in 80nm, which is a half-node, so called "optical shrink" of 90 nanometres, there is less and less doubt that this baby packs at least half a billion transistors. People forget that the number of metal layers is also increasing, so allowing additional hundred million or so transistors to be packed in.
[img]http://www.short-media.com/images/newsimages/2006/November/R580 600 comp.jpg[/img]
Industry news sets the R600 launch for January 20th, 2007
Source: The Inquirer
0
Comments
geeks.:type: *boom*tish*
i know i know. i *had* to.
http://www.theinquirer.net/default.aspx?article=35708 Though, far from specs, its a very good link to see whats to come.
cheers.
-isi
oh well
you stole my post ! :P rotflol
lol thrax is RITE
I'm <i>really</i> internet lazy, but I could go through the last two years of the Inq and pick out dozens upon dozens upon dozens of articles wherein the Inq accurately called it long before anyone else.
The Inq is rarely wrong, and when it is, it says it is. Respect to the Inq.
Throughout my computer expierience i have no idea where people have gotten the wrong idea of theinq. I have been on the computer scene since 2002 and have read their articles over and over, and yet RARELY continue to find errors.
If I can find a link to a different site that hosts the same story I usually use that instead. My reason for doing so is that most of the articles the Inquirer puts out are so badly written I'm embarrassed to link to them.
If anyone has found a novel way to slaughter the English language, they have.
I think therein lies their bad rep. Really good research is killed by really bad writing.
Plus, Fuad Absdxzxvbxchfgxzovic is a lazy, fat bastard, and we all know that.
As far as the writing goes, i'm surprised many people are not able to see past the ubiquious and opaque writing style. Even as a first reader, i was able to read it transparently without much effort and actually found the articles original in terms of written context. I guess its just a matter of preference i suppose, you know admiring the qualitie(s) of being politically correct and all. Besides, i'm sure there are alot of others that totally disagree that their writing style [strike]or blogging style by Thrax[/strike] affects the overall quality of the article, as we geeks possess an open mind anyway.