More Pipelines or Higher Core Frequency ???

[pr0pensity]

More features.

 

[Elios]

iirc some site did a clock for clock test setting a 7800GTX to 16pipes and down clocking a X1800XT to the GTX clocks and the 7800 beat it

so in theory if NV did a die srink and clocked up a 24pipe card to the speeds of the X1800...

i have had both ATi and nVidia cards and the IQ seems the same to me though NV seems to have a some issues with moire with transparent textures

 

[Trimlock]

iirc some site did a clock for clock test setting a 7800GTX to 16pipes and down clocking a X1800XT to the GTX clocks and the 7800 beat it

so in theory if NV did a die srink and clocked up a 24pipe card to the speeds of the X1800...

i have had both ATi and nVidia cards and the IQ seems the same to me though NV seems to have a some issues with moire with transparent textures

the fact is, the ATI card isn't ment to be used with that down clock, and the 7800 isn't ment to be used with 16PP, it just proves you can't theorize performance anymore by whats on paper, if nvidia does do a die shrink and up the clocks (thats to say it can, we still don't know) if it sees a nominal performance increase

reason being who cares at what clock, or how many pipes the cards have and compares to each other, its how it performs in real life that matters

 

[Patman]

ATi for the insane overclocking.

agreed, ATI is Die[H]ard

 

[Sc4freak]

I didn't read the entire thread, but here is my take.

Consider an Nvidia 7800GTX. 430mhz, 24 pixel pipelines. That gives it a peak fillrate of 10320 MTexels/sec. If we reduce the pipelines to 16, and raise the clockspeed to 645 mhz, then the peak fillrate is still 10320 MTexels/sec.

BUT, with 645mhz, the GTX's 8 vertex shader units now work at 150% of their original speed at 430mhz. Its 16 ROPs also run 50% faster, due to higher clockspeed.

All while maintaining the same peak pixel fillrate.

 

[Elios]

the fact is, the ATI card isn't ment to be used with that down clock, and the 7800 isn't ment to be used with 16PP, it just proves you can't theorize performance anymore by whats on paper, if nvidia does do a die shrink and up the clocks (thats to say it can, we still don't know) if it sees a nominal performance increase

reason being who cares at what clock, or how many pipes the cards have and compares to each other, its how it performs in real life that matters

well with both cards running the same clock and pipes the 7800 was still faster and you cant say "well it wasnt ment to run slow" the point is the 7800 is getting more done

 

[robberbaron]

the fact is, the ATI card isn't ment to be used with that down clock, and the 7800 isn't ment to be used with 16PP, it just proves you can't theorize performance anymore by whats on paper, if nvidia does do a die shrink and up the clocks (thats to say it can, we still don't know) if it sees a nominal performance increase

reason being who cares at what clock, or how many pipes the cards have and compares to each other, its how it performs in real life that matters

Ehh, you still can to a pretty big degree. The fillrate of the X1800XT is still about the same as the 7800GTX, and they have the same number of vertex pipes. What a surprise, they perform about the same, give or take in some areas.

 

[Trimlock]

well with both cards running the same clock and pipes the 7800 was still faster and you cant say "well it wasnt ment to run slow" the point is the 7800 is getting more done

no it was ment to run the way it is ment to run, its not ment to run slow, and neither is the ATI card

and robber, i agree, there isn't any hidden agenda between the two cards =/

 

[Elios]

no it was ment to run the way it is ment to run, its not ment to run slow, and neither is the ATI card

and robber, i agree, there isn't any hidden agenda between the two cards =/

that same can be said for the other way then 7800s and X1800s arnt ment to be run faster ?
your logic is wrong

 

[noobman]

When I first saw ATi new line-up, i was unsettled by the fact that they kept the same number of pipelines than the previous gen. I have not seen this happen from one gen to another since the time when pipelines weren´t an issue...

Now seeing that both cards perform very competitive with one another which company do you think made the rigth choice? nVidia for increasing its number of pipelines while keeping speeds the same, or ATi for increasing clock frequencies and keeping the same number of pipelines?

This is not an ATi Vs nVidia post.... though it might seem like it... is more like the title says... which do you prefer, more pipes or higher clocks...

Well, we should probably edit your poll choices then One is labeled ATI, the other Nvidia - DL

It all depends on the game really. It's best to strike a good balance... you don't want a slower card with excess pipelines(XGI Volario, that was ONE of it's MANY problems), but on the other hand a super-fast card lacking bandwidth (GeForce FX 5800, again one of it's MANY problems) is no good either.

This will become a lot simpler once both companies start using unified pipelines (which can do both vertex and pixel shaders and will adapt to suit the scene). They're less effective, but provide better bandwidth in the end.


Also remember that the only reason the ATi cards don't have 32 pipes right now is because they couldn't manufacture them (basically, 3 out of every 100 chips they made qualified for use). Part of the problem was them using the 90nm processor.... which indicated that they planned on having higher clock fequencies than the nVidia cards in the first place.