The Rumor Roundup

Doomtrooper said:
caboosemoose said:
These "sources" are said to be "familiar" with the roadmaps of both companies, but the overall tone of this blurb is too reminiscent of the rumour mill.

Yeah, that story ir rubbish.

FYI Digitimes is rarely wrong, many investors rely heavily on them..look at their past track record. Digitimes is not the inquirer.

Digitimes may be the finest website in CHristendom (and outside christendom too, in the circumstances). They may never have been wrong before. They are this time. Exactly how they are wrong is another matter.
 
The Rumor Roundup

My best friend has told me that my wife is deceiving me!

Not too bothered so long as she doesn't bother me.

The point is .... it's all rumour. We had it before & it was an annoying experience. In my case it led to frustration & a serious let-down. I would suggest that those people with "a little Birdie" be a little cynical & consider the possibilty that they are being used.

Oh yeah, the wife is deceiving me by saying we can't afford an AMD64 system. ;)
 
aaronspink said:
A little fuel for the fire though... Has anyone thought of the possibility of the R420 having multiple clock domains? If they are reusing pretty much intact the architecture of the R300 for the pipelines, then that means they would have had on the order of 18 months to do custom circuit work. What if the pixel pipes are double pumped? That would certainly be extreme in my opinion, the limitation on pixel output to memory is 12 pixels per cycle.

Imagine an 8 pipe R420 running at 500-600 Mhz able to run its pixel pipes at 1000-1200 Mhz with a 12 pixel output per cycle. Should easily be able to get 2-3x the performance of R360. It would also explain the rumors of 700-800 Mhz GDDR3 for it as well.

Aaron Spink
speaking for myself
Sounds most plausible to me and, IIRC, If you add in the extra transistors for the rumored additional VS units you may just have the recipe for he r420.
 
nelg said:
aaronspink said:
A little fuel for the fire though... Has anyone thought of the possibility of the R420 having multiple clock domains? If they are reusing pretty much intact the architecture of the R300 for the pipelines, then that means they would have had on the order of 18 months to do custom circuit work. What if the pixel pipes are double pumped? That would certainly be extreme in my opinion, the limitation on pixel output to memory is 12 pixels per cycle.

Imagine an 8 pipe R420 running at 500-600 Mhz able to run its pixel pipes at 1000-1200 Mhz with a 12 pixel output per cycle. Should easily be able to get 2-3x the performance of R360. It would also explain the rumors of 700-800 Mhz GDDR3 for it as well.

Aaron Spink
speaking for myself
Sounds most plausible to me and, IIRC, If you add in the extra transistors for the rumored additional VS units you may just have the recipe for he r420.

* Natoma goes to b3d-babelfish to understand wtf this means.....

If you please. :)
 
2senile said:
Oh yeah, the wife is deceiving me by saying we can't afford an AMD64 system. ;)

This is why I'm getting married shortly after Socket 939 Athlon64 systems become available to buy. :D
 
digitalwanderer said:
2senile said:
Oh yeah, the wife is deceiving me by saying we can't afford an AMD64 system. ;)

Hey, mine too! :oops:

It's a conspiracy I tell ya... :(

Damn them all! Why don't the female of the species understand our "needs"?

Semi-serious Digi', I posted this in my fav' internet hang-out; ".... as they have already stated that their next parts will have lower margins than they are currently achieving.". It was from page two of the Cebit report by "wavey". What makes me wonder is if the lower margins are because they have an expensive killer card or if they have to price it lower because the competition has more. (features/speed)
 
Finances?????????????????

Is that money?

Last time I handled the finances it was the barter system. I got an Oxen for three bags of wheat. :(
 
2senile said:
Semi-serious Digi', I posted this in my fav' internet hang-out; ".... as they have already stated that their next parts will have lower margins than they are currently achieving.". It was from page two of the Cebit report by "wavey". What makes me wonder is if the lower margins are because they have an expensive killer card or if they have to price it lower because the competition has more. (features/speed)

Lower margins. That's also the reason I've been hearing for ATi delaying the R420's release until the nV40, they're making more money off of the R3xx line than they will off the R420 line-up. (But they'll still make money on the R420 line-up, just not as much of a profit as on the R3xx line-up.)
 
Natoma said:
Psst... That's why it's better when you handle the finances. I should know. ;)

NO WAY! :oops:

You have no clue how badly I suck with money, I want as little access to me families savings as possible! :?

It ain't really a proud point, but I do know my limitations. ;)
 
nelg said:
aaronspink said:
A little fuel for the fire though... Has anyone thought of the possibility of the R420 having multiple clock domains? If they are reusing pretty much intact the architecture of the R300 for the pipelines, then that means they would have had on the order of 18 months to do custom circuit work. What if the pixel pipes are double pumped? That would certainly be extreme in my opinion, the limitation on pixel output to memory is 12 pixels per cycle.

Imagine an 8 pipe R420 running at 500-600 Mhz able to run its pixel pipes at 1000-1200 Mhz with a 12 pixel output per cycle. Should easily be able to get 2-3x the performance of R360. It would also explain the rumors of 700-800 Mhz GDDR3 for it as well.

Aaron Spink
speaking for myself
Sounds most plausible to me and, IIRC, If you add in the extra transistors for the rumored additional VS units you may just have the recipe for he r420.

I remember talk of a double pumped design before around about the time ATi announced that it was using Fast14 Technology. Now I don't know if there is any sort of inter connect between the ideas I can't remember the exact discussion but the double pumped solution was supposedly an unrealistic expectation. Something about them 8 extreme pipelines though makes me think that possibly ATi has done that. Dear god, I can hardly stand all this speculation. I wish that ATi would simply let the specs slip out into the real world.

EDIT: For reference to the Fast14 technology agreement here is a quote from the PR announcement.

"We selected Intrinsity after determining that Fast14 Technology can deliver up to four times the performance per silicon dollar when compared with standard design approaches."

Fast14 Technology provides a sustainable advantage in design technology, enabling chip designers to be more productive and deliver faster circuits by automating the design of high-speed logic. Unlike desktop processors that achieve multi-GHz speeds with time-consuming and expensive handcrafted custom circuits, Fast14 Technology achieves desktop processor speeds with short design cycles and small design teams.

http://biz.yahoo.com/bw/040205/55824_1.html

I guess the question is whether ATi intends to use this technology with the R420 or products further down the road.
 
Well Sabastian, to continue with this line of speculation, perhaps ATI wanted to go the double pump route and found that it is very difficult. This is turn lead them to try this technique on a proven design (which would be consistent with ATI’s modus operandi) hence the cancelling of the original R400. This of course would make them much more appreciative of the design tools that FAST14 developed.

Again I state that Arron's theory is a very sound one (IMHO).
 
nelg said:
Well Sabastian, to continue with this line of speculation, perhaps ATI wanted to go the double pump route and found that it is very difficult. This is turn lead them to try this technique on a proven design (which would be consistent with ATI’s modus operandi) hence the cancelling of the original R400. This of course would make them much more appreciative of the design tools that FAST14 developed.

Again I state that Arron's theory is a very sound one (IMHO).

heh, I have my own theory regarding the R400 design. I think ATi pulled it for use in the xbox2 tweaked up of course. Instead of attempting to design a chip from scratch for MS they decided to use what they had. Fast14 might be for that project as well. The R420 contingency plan came because of the decision to use the R400 for the next xbox...hows that for speculation? ;)

I can not say one way or the other about what he is suggesting only that some thought that the idea was highly speculative and not likely. Surely it would be an impressive solution.
 
Sabastian said:
nelg said:
Well Sabastian, to continue with this line of speculation, perhaps ATI wanted to go the double pump route and found that it is very difficult. This is turn lead them to try this technique on a proven design (which would be consistent with ATI’s modus operandi) hence the cancelling of the original R400. This of course would make them much more appreciative of the design tools that FAST14 developed.

Again I state that Arron's theory is a very sound one (IMHO).

heh, I have my own theory regarding the R400 design. I think ATi pulled it for use in the xbox2 tweaked up of course. Instead of attempting to design a chip from scratch for MS they decided to use what they had. Fast14 might be for that project as well. The R420 contingency plan came because of the decision to use the R400 for the next xbox...hows that for speculation? ;)

I can not say one way or the other about what he is suggesting only that some thought that the idea was highly speculative and not likely. Surely it would be an impressive solution.

the only problem with that theory is its pretty much been said that Xbox2 would be based on the R500
 
Back
Top