David Kirk Interview

[ The Kirk person ] story already a little is complicated. The Vertex Shader of the GeForce FX consists of the plural floating point arithmetic processors of multi-threading correspondence. In other words, the Vertex Shader is not the case that it is separated into the pipeline completely. Until recently, pipeline architecture was changed largely. Therefore, method of counting that is not possible what is the Vertex Shader.

I think we can say that Digit Life was incorrect on the 3 vertex shaders statement!
 
DaveBaumann said:
I think we can say that Digit Life was incorrect on the 3 vertex shaders statement!

Digit Life's "Guesstimation Preview" seemed far too one sided to be believable, I thought.

Anyway, ''yield aegagropila badness passes'' is properly translated:

"[The Kirk person] had a mild attack of sour balls which is sure to end soon." (Couldn't really tell if this was in context.)

Nor did I see, "the Jim dead is he" anywhere, which is surprising (trans: "He's dead, Jim", or "Jim is dead," not really sure.)
 
Well Chalnoth, maybe you could get your Japanese roommate to give us a proper translation to that interview??

Would be nice to really know what is being said.

US
 
"I think we can say that Digit Life was incorrect on the 3 vertex shaders statement"

but still. doesn't GFFX have the roughly the same vertex performance as if it had 3 vertex shaders? regardless, GFFX has LESS vertex performance per MHz compared to R300.
 
andypski said:
I particularly like the translated phrase :

'yield aegagropila badness passes'

It makes it sound more like he's discussing his hemorrhoids than graphics hardware. :LOL:

As far as I can discover aegagropila is not even a word, so I don't know where the translator got it from...

hahahahahahahahahahahahaha, thats was funny?
 
A curious phenomenon is associated with a certain type of Cladophora (algae) that forms free-living spheres, filled with water, mud or gas released from photosynthesis. These ball forms, called Aegagropila, may be found in both freshwater lakes and nearshore marine environments, and vary from a diameter of a few millimeters to the size of a human head!

? Still makes no sense, oh well, better that "sour balls"

-Simon
 
Mr. Kirk's important statement:

GeForce FX's displacement mapping function is "render to vertex buffer"
and "displace by pixel shader".

I don't know wheter special hardware support for this process.
Mr. Andreas Kugler(now NVIDIA) present paper concerned this process
several years ago.
http://www.eg.org/EG/CGF/volume20/issue1/cgf471.html
This paper's method requires special hardware support.
 
LMFAO LITERALLY!!!
I was laughing so hard I started crying. (literally)

It is the very deep pipeline where [ the Kirk person ] so is.

In other words, the Vertex Shader is not the case that it is separated into the pipeline completely.

[ The Kirk person ] so is. Fixed function & T L

Also the ?????????? is similar

Whether or not [ the Kirk person ] pipeline change was done?

As for 0.13 mu m process

[ The Kirk person ] as for the reason it is not understood, but with the GPU, as for copper wiring the big difference was born. 0.13 With mu m as for copper wiring, with 0.18 mu m process of aluminum wiring compared to as many as 140% or more circuit speed becomes high speed. As for advantage of copper wiring not only acceleration, low there is also a electric power consumption. The to high clock converting, because () resistance of wiring does not increase that much.

*puts on Yoda's voice* made narrow an advantage, even when designing there is.

When just this it tries probably to make the powerful pipeline, you could not think 0.15 mu m process.

I the R300 (RADEON 9700) call the last product of the past generation.

0.15 With mu m, (the die/di) becomes too large, heat generation becomes too many, the clock be too low, yield aegagropila badness passes. The negative element is too multi.

[ The Kirk person ] so is. Heat had become big restriction.

...the technology itself of DDR II. It became good timing.
 
Back
Top