X800Pro,, 16 pipelines

[Soulburner]

Originally posted by StormPC
How could it do that if CPU doesn't matter?

Who said CPU doesn't matter?

Where are you going with this?

[StormPC]

Originally posted by Soulburner
Who said CPU doesn't matter?

Where are you going with this?


"Mother Nature isn't really affected by CPU at all."

"much lower than it should be, not even matching his score at default clocks."

"That is ALL GPU and shader power there."

You don't see how these don't exactly jive with your statement:

"Maybe because the EE outscores the A64 in that benchmark."


With all due respect, it takes 4 game tests to score well in 2k3, not 1. Whatever...we've beaten this horse enough I guess.

Maybe I need a pencil mod???

[Soulburner]

Ok then maybe you didn't know this...which is fine.

GT1 is a DX7 test, which relies heavily on CPU speed much as 3DMark2001 does.

GT2-3 your CPU has a very minimal impact on.

GT4 your CPU has almost no impact on.

Macci knows his stuff, he wouldn't say something like that unless he really meant it (and yes even after taking into account your CPU).

[Mini]

Originally posted by StormPC
With all due respect, it takes 4 game tests to score well in 2k3, not 1. Whatever...we've beaten this horse enough I guess.

Maybe I need a pencil mod???

There is always a place where programs are CPU power laking and GPU power laking...

In 3dmark03 with R3XX series it was GPU laking... Now with R4XX its CPU laking...

[StormPC]

Originally posted by Soulburner
Ok then maybe you didn't know this...which is fine.

GT1 is a DX7 test, which relies heavily on CPU speed much as 3DMark2001 does.

GT2-3 your CPU has a very minimal impact on.

GT4 your CPU has almost no impact on.

Macci knows his stuff, he wouldn't say something like that unless he really meant it (and yes even after taking into account your CPU).

I know that but you have to realize that due to differences in the P4 and A64 architectures sometimes less is more.

I know what it is. He's using modified DNA CATs and I'm using stock drivers.

Anyway, what he gets is not very meaningful because he is the tweaking god!

FOR GOD's SAKE, I JUST GOT THE CARD LAST NIGHT!!!

Those were my first runs! Forgive me...


MINI:

Good point dude. CPU makes more of a difference with the new cards than it used to because they ARE now the bottleneck!:thumbsup:

[Marquzz]

The mod failed on a Club3D I have atm. Some data for experience:

[SuperDude]

marquzz, try it again. it looks like your compound is touching the dot in the middle just a bit. remove it and see if you can get a perfect line

[Hallowed]

Soul = patience++

[Boogotop]

Originally posted by Marquzz
The mod failed on a Club3D I have atm. Some data for experience:

i got the same stepping on my MSI and doesn't work

[IronYuppie]

Could someone please comment on what's posted on the front page:

ATI responds quickly
I spoke with ATI and confirmed that the mod works but the boards will not survive very long. We should expect to see reports of failed attempts and boards that die. . .


I imagine this could only be because of the increased clock rates (which seems always the official line with clocks), but could the procedure of the pipe-mod, even assuming it's done correctly, shorten lifespan in your opinion?

[SuperDude]

probably just a scare tactic so that people will stop modding and getting more for they pay for

[StormPC]

Originally posted by SuperDude
probably just a scare tactic so that people will stop modding and getting more for they pay for

For sure. I think they also want to discourage the mod so they don't have to deal with all the RMAs from people who messed up while doing it.

[r3b0rN]

lol stormpc this is why i wouldnt want to be the first to get a beasty card as it just came out (in this case the x800xt) people expect you to be getting the best results on your first run, when you havnt even formatted yet and your just mucking around with it. anyways, have u found out how to switch to 1T? or are you going to get another 1 later?

[Soulburner]

I don't think anyone expected anything from him, macci was only making a valid point about his score being low. Considering his card is overclocked and is behind a stock X800Pro in performance, it may be a cause for concern.

[StormPC]

Originally posted by r3b0rN
lol stormpc this is why i wouldnt want to be the first to get a beasty card as it just came out (in this case the x800xt) people expect you to be getting the best results on your first run, when you havnt even formatted yet and your just mucking around with it. anyways, have u found out how to switch to 1T? or are you going to get another 1 later?

The option is in the BIOS when the 3700+ is installed but not the 3400+, I'm guessing because it defaults to 1T. The problem is that it doesn't actually work. Have not determined whether it's a bug or whether my hardware is faulty.

Soulburner,

I thought I explained the low score pretty well. Macci was doing his Jason impression, that's all. Can we move on yet?

[Naser]

I think its all in the CODE (3rd Line), example if lets say a card with code GA8999.1 was successful then ALL cards with the same code should work, it has nothing to do with the last digit example GAXXXX.1 nor the BBA nor the brand itself.Ofcourse this isnt 100% right but im willing to bet on it.

I got the Sapphire X800PRO & i did the mod at least 5 times to make sure but still with no luck, the steppings on the card are as follows :

215RAACGA11F
GB7293.1
0420AA
you can also see its week 20 of 2004 which is one of the newest cores.

Now i went ahead & bought another card made by GeCUBE & when i opened it, it showed this steppings:
215RAACGA11F
GA8999.1
0416AA


I havent got to modding it yet but if my theory is right it should work.,since most people with the same code were successfull.
I'll let you guys know how it goes hopefully by tonight.

EDIT: Added images of both cards.

[macci]

Interesting info there Naser. Let us know how the CeCube card works

StormPC, sorry the post I made was a bit evil

BUT the default Nature scores I mentioned there were both ran on my everyday PC which has a normal P4C800-E + 3.0C NW CPU combo in it. So it shouldn't be the reason for the fairly high difference we have in the Nature fps.
ATI-Drivers/D3D settings prolly make a difference, etc but it could be a chipset/agp/driver -problem too. 70 vs. 87 is quite a difference considering that we are basically useing the same hardware (and default clocks).
If you get the Nature FPS tweaked to the optimal level it'll give quite a nice boost for the score too

[Bastieeeh]

OT:
@ Marquzz and Naser: What digicams are you guys using? These are very detailed shots! I am about to buy a digicam myself and am looking for high quality stuff. So pls be kind and tell me ... ;-)

[r3b0rN]

Originally posted by macci
Interesting info there Naser. Let us know how the CeCube card works

StormPC, sorry the post I made was a bit evil

BUT the default Nature scores I mentioned there were both ran on my everyday PC which has a normal P4C800-E + 3.0C NW CPU combo in it. So it shouldn't be the reason for the fairly high difference we have in the Nature fps.
ATI-Drivers/D3D settings prolly make a difference, etc but it could be a chipset/agp/driver -problem too. 70 vs. 87 is quite a difference considering that we are basically useing the same hardware (and default clocks).
If you get the Nature FPS tweaked to the optimal level it'll give quite a nice boost for the score too

any computer owned by you automatically performs 10% better than any1 else with the exact same hardware

[Marquzz]

Originally posted by SuperDude
marquzz, try it again. it looks like your compound is touching the dot in the middle just a bit. remove it and see if you can get a perfect line

Yes, already did it 3 times, as you can see on this picture it's nothing wrong with the mod itself







Using a Nikon 775 for those who wonder, Nikon has one of the best makros there is.

[Naser]

Originally posted by Bastieeeh
OT:
@ Marquzz and Naser: What digicams are you guys using? These are very detailed shots! I am about to buy a digicam myself and am looking for high quality stuff. So pls be kind and tell me ... ;-)

The first picture was using the Fujifilm FinePix 6800Z

the next one is with the SONY DSC-F828 which i dont not much about but definatly much better then the finepix.

[impu|se]

IMO, Nasers conclusion is the only one that can be true. Its either what Naser said, or that it is just pure luck to get a card that mods, which i dont believe.

Originally posted by r3b0rN
any computer owned by you automatically performs 10% better than any1 else with the exact same hardware

Works on the same principle as a TypeR sticker

[BioBlade]

here is my mod :




and the card cooling modded but not tested, because my chiller is not working..

[LilGator]

Someone over at overclockers.co.uk thinks this might work, as it will connect ALL pipes together... anyone that has failed before want to give it a shot ?

Naser, Chris_UKs card was a:

GA8999.1
0416AA

[IronYuppie]

Originally posted by SuperDude
probably just a scare tactic so that people will stop modding and getting more for they pay for

Yeah thats what I thought too. I so hope this mod works on my sapphire!

Almost here! It's on the fedex truck right now