• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Mod your X800Pro into an XT NOW !!! what are you waiting for!!!!

Originally posted by: Avalon
Let's all post a bunch of letters. Why? Because it's...

P
O
I
N
T
L
E
S
S

I dont get it. Why are ppl posting random letters? 😕

Read bits of the linked boards. Firstly - Overclockers UK - same place i usually buy my hardware - had no idea they had a forum.

Secondly - seems like alotta work to turn a PRO into an XT. Well, for me maybe as i have a phobia when it comes to any type of DIY job. So i guess then that the x800 PROs can have their pipes unlocked, but not thought softmods - although im not entirely convinced yet.

Oh well - makes little difference to me as i dont plan to buy a new gfx card anytime soon, but kudos to the UK boys there 🙂
 
They're posting random letters that spell the word REPOST, because this thread is a REPOST (it's already been posted about)
 
Originally posted by: Lonyo
They're posting random letters that spell the word REPOST, because this thread is a REPOST (it's already been posted about)

Ohhh. Well - the T was missing. In any case, would have been better for someone to have just said that. 🙂
 
Ya know what I find odd? 33% more pipelines = 10-15% more performance... in their 3DMark2003 tests at least. Looks like the rest of the card (memory system) may not be strong enough to handle all 16 pipelines.
 
Originally posted by: Jeff7181
Ya know what I find odd? 33% more pipelines = 10-15% more performance... in their 3DMark2003 tests at least. Looks like the rest of the card (memory system) may not be strong enough to handle all 16 pipelines.

That kinda crossed my mind also. I would have expected to have seen much higher increases in their 3dmark scores.
 
seems to be risky. some boards didn't work. some defects in ati demos. overclocking required to reach XT specs. overclocking varies. results in 3dmark don't seem too hot.
 
Originally posted by: CaiNaM
why would you find that odd? it's not like the XT is 30% faster than the PRO....

Well supposedly the x800 is a pixel shader powerhouse... so I figure in a game like Far Cry where pixel shader performance is critical, a >30% increase in pixel pipelines (not to mention clock speed) would provide more than 10% more performance. Unless of course, as I said, it's being limited by memory bandwidth.
 
Originally posted by: Jeff7181
Originally posted by: CaiNaM
why would you find that odd? it's not like the XT is 30% faster than the PRO....

Well supposedly the x800 is a pixel shader powerhouse... so I figure in a game like Far Cry where pixel shader performance is critical, a >30% increase in pixel pipelines (not to mention clock speed) would provide more than 10% more performance. Unless of course, as I said, it's being limited by memory bandwidth.

far cry? there are no fc benchmarks.. the only point of comparison we have is (and the one you referenced) is 3dmark03.

my point was that the higher clocked xt w/ 4 additional pipes isn't 30% faster, so why would a modded pro be any faster?

6800UE: 12759
X800XTPE: 11895
6800U: 11868
x800PRO: 9639

http://www.firingsquad.com/hardware/geforce_6800_ultra_extreme/page5.asp

my PRO w/o MOD: 10963
 
Originally posted by: CaiNaM
Originally posted by: Jeff7181
Originally posted by: CaiNaM
why would you find that odd? it's not like the XT is 30% faster than the PRO....

Well supposedly the x800 is a pixel shader powerhouse... so I figure in a game like Far Cry where pixel shader performance is critical, a >30% increase in pixel pipelines (not to mention clock speed) would provide more than 10% more performance. Unless of course, as I said, it's being limited by memory bandwidth.

far cry? there are no fc benchmarks.. the only point of comparison we have is (and the one you referenced) is 3dmark03.

my point was that the higher clocked xt w/ 4 additional pipes isn't 30% faster, so why would a modded pro be any faster?

6800UE: 12759
X800XTPE: 11895
6800U: 11868
x800PRO: 9639

http://www.firingsquad.com/hardware/geforce_6800_ultra_extreme/page5.asp

my PRO w/o MOD: 10963

For the record I never said there were any Far Cry benchmarks 🙂

Anyway... for the third time... maybe there isn't enough memory bandwidth available for 16 pipelines whether it's a modded Pro or an XTPE... would be interesting to see what effect increasing JUST the memory speed has
 
Anyway... for the third time... maybe there isn't enough memory bandwidth available for 16 pipelines whether it's a modded Pro or an XTPE... would be interesting to see what effect increasing JUST the memory speed has
That's probably why the XT comes with much faster memory than the Pro. I agree with you though. The softmod alone only seems to be yielding 15% performance, to get the full 25-30% you'd probably need a nice memory overclock.
 
Originally posted by: GeneralGrievous
Anyway... for the third time... maybe there isn't enough memory bandwidth available for 16 pipelines whether it's a modded Pro or an XTPE... would be interesting to see what effect increasing JUST the memory speed has
That's probably why the XT comes with much faster memory than the Pro. I agree with you though. The softmod alone only seems to be yielding 15% performance, to get the full 25-30% you'd probably need a nice memory overclock.

doesn't matter the XT comes with "faster" memory.. pro memory can still attain xt speeds.

the problem is the driver is not fully optimized, only yielding somewhere around 70% of r420's theoretical bandwidth.

ati has stated thru a more efficient compiler, etc. they should be able to attain another 10-15% increase in performance.
 
I expect once this card is readily available I'll be seeing a lot of them in the refurb section 😉
 
Originally posted by: GeneralGrievous
Anyway... for the third time... maybe there isn't enough memory bandwidth available for 16 pipelines whether it's a modded Pro or an XTPE... would be interesting to see what effect increasing JUST the memory speed has
That's probably why the XT comes with much faster memory than the Pro. I agree with you though. The softmod alone only seems to be yielding 15% performance, to get the full 25-30% you'd probably need a nice memory overclock.

its not a softmod man. do you not know the difference?



Originally posted by: DAPUNISHER
I expect once this card is readily available I'll be seeing a lot of them in the refurb section 😉


yes indeed all from asses who bought the card and "it didnt work"
 
Originally posted by: CaiNaM
why would you find that odd? it's not like the XT is 30% faster than the PRO....

You are right I actually tried to estimate it and it's slightly below 30%

Ok lets see, since most people buy these top end cards to play the latest games with great eye candy I will compare only 1600x1200 4AA/8AF performance, otherwise why buy these cards?

Quake 3 - 31.2%
UT2k4 - 16.5%
Call of Duty - 15.7%
Far Cry - 35.7%
Halo - 36.3%
SplinterCell - 28-29.3%
Tomb Raider:AOD - 21.8-27.3%
Homeworld 2 - 35.4%
EVE: The Second Genesis - 48%
Jedi Knight: Jedy Academy - 26.3%

3dMark03 - 20.3%
BUT in this X800xt got 31.2% faster in game 2, 26.4% in game 3

Well I wouldn't say the performance is 30% across the board and would roughly say its between 15-30% faster and definately not 0-15% but more on the upper half of 15-30% for x800xt over x800pro, with greatest increases in the most intense games (which is most important). I think it would be fair to say for the most part, X800XT is ~25%-30% (by taking the rough average) faster.
 
Back
Top