8800gts vs 2900 pro all tests made by me

[ccleorina]

Wow.... What a WAR Nvidia VS ATI.... All i can say is my ATI HD2900XT CrossFire can kill 8800GTS 640 in SLI...

[Wile E]

Holy crap. Can I get a thread summary? lol

[AphexDreamer]

Originally Posted by Wile E

Holy crap. Can I get a thread summary? lol

Ok, you asked for it. So basicly the dude posted up some benchies comparing the 8800gts with 2900Pro. People were commenting on it with postive thoughts and then other people were analyzing it further and thought that the CPU was just not cutting it for the overclocked HD2900Pro. People were like it can do better and overclock higher if only it had a higher clocked CPU and then others were like yall full of shit and know nothing so shut up. But the people were brave and defended what they thought was right and so did the others, in result the storm slowly brewed and now has been unleashed. I think that some what somes it up, I'll come back and edit this later when its not 6:00 A.M.

[ccleorina]

Holy crap... This thread is out of the topic...... Can any one here stop it.... HD2900 did need higher CPU clock to OC higher clock speed... I did get good OC on my HD2900 with OC my CPU.... What the point WAR here....? Try your self.... I can sure about it....

U are right AphexDreamer

[Tatty_One]

There is no need to start a flame war here between the 2 of you, we are all entitled to our opinions, in this case one thinks that the futuremark score is a single card and legitimate, the other does not, I happen to go with Trt on this one, not because of anything else but the following FACTS that I will post, the supposed futuremark link you posted, you beleive to be a single card setup and had the core clock running at 720Mhz with a shader clock of 1832, on a Q6600 at 3.6Gig, now I do not beleive that a 14000+ score can be acheived on a 320MB GTS single card setup for these reasons........

The 8800GTS unofficial world record for 3D Mark 2006 stands (or stood), as at September 2007 by a guy known as "Youngpro" from Australia at 15,053 and that was with a 640MB 8800GTS on a Quad extreme QX6700 at 4.550Gig! the 8800GTS was on liquid nitro I beleive and clocked 803 Mhz on the core! by my maths that made the shaders come in at around 1860, now compare those clock speeds and CPU setup with the link you provided and that cannot be a single card score.....simple as that. The difference in the Quad core speeds alone (3.6Gig in your link to 4.550Gig in mine) would practically account for the records 800 points increase let alone the extra 83Mhz on the core of the GTS, never mind the fact that it was a 640.

Post 41...... http://www.xtremesystems.org/forums/...=163000&page=2

Just my thoughts, of course I could be wrong, incidentially, the world record before that was just 130 points higher than the futermark link you posted, that was on a 320MB GTS but with a core speed of 836Mhz! and a Q6700 at 4Gig, you can see from those specs alone in comparison to your fututremark link that it seems at least unlikely that a 720Mhz core 320 with just 3.6Gig is going to get near that? Like I said, we are all entitled to our opinions but there is no need to start with insults because another does not agree, I liked your origional review.....thank you!

[cefurkan]

14 k sli skor is nothing

he get 14 k with single card and it is possible

if u dont know these cards dont make such comments

i can take 12 k with e4300
with 4k quad core it is so easy to hit 14 k with last drivers

stiil such ignorants claims that u cant take 14k with single card

lol

here is the single card scores of hwbot

http://www.hwbot.org/searchResults.d...inTotalPoints=

max is 15601

and here is sli scores

http://www.hwbot.org/searchResults.d...inTotalPoints=


max 20921

[trt740]

Originally Posted by cefurkan

i think u are retarded

u called 14k score sli

but it was single card score

than i put real sli score

19 k

but since u are a retard

u cant understand these scores

Please I preferr specials needs.

[Mussels]

Originally Posted by cefurkan

i think u are retarded

u called 14k score sli

but it was single card score

than i put real sli score

19 k

but since u are a retard

u cant understand these scores

in my PC a GTX gets 19-19.5K, while a GTS 640 gets around 13.5K

so yes, i beleive your single card score.

[Tatty_One]

Firstly, I will make such comments as I like (remember you are the one who has insulted other members of this forum, not me ), thats what these forums are about... opinion, you may feel you are in a position to tell people what they can or cant say......but your not!, secondly, I was actually referring to the link you made to futuremark and the score there, unless VRZone and Australia's Atomic PC are lying about just over a 15000 world record, how do you explain how that fairly basic setup you linked managed to get fairly close to an "extreme" setup that cracks the world record? We could argue all day about 14000, I dont care about that, it's clear that with extreme systems, 15000 scores can be acheived, what you posted was nowhere near those systems, all I am concerned with is the accuracy of the link you provided is all, all along I have not questioned the "extreme" potential of these cards, damn many cards have that kind of potential with a 4.55Gig Quadcore backing them up, some serious over volting and liquid nitrogen to back them up, a Q6600 at 3.6Gig and a 320MB GTS at 720core is not one of those extreme systems.

Look, bottom line is this, people agree and disagree, have different opinions, so what? no biggy.....live with it. I am happy that you are so confident that you can crack more than 12000 with your rig (I can with my rig) but thats not the point, now if you said you could crack 13000 I would sit up and listen because that would mean that with a Q6600 @ 3.6Gig you might get close to 14000. I am off to lie down in a dark room

[Tatty_One]

Originally Posted by Mussels

in my PC a GTX gets 19-19.5K, while a GTS 640 gets around 13.5K

so yes, i beleive your single card score.

19-19.5? we are talking single cards here, he is talking a single 320mb getting over 14000 from a Q6600 @ 3.6Gig, I am not saying that 14000+ scores cant be acheived, they clearly can, just dont think an "everyday" system they are, damn check the 3D Mark 2006 thread, there are GTX's on quads getting nowhere near 14000. In fact the 2nd highest score is a GTX, well overclocked on a Q6600 at 3.8Gig and thats got just over 14000!!! i rest my case.

I will say no more on the subject.....I am getting bored now

Edit: Taken from HWbot, 14,197, these are the rig specs:

Processor: Core 2 Q6600 (2.4Gh... @ 3800mhz no image
Videocard: GeForce 8800 GTS 32... @ 783/1026mhz
Global Rank: 261st - 5.3 points
Hardware Rank: 7th GeForce 8800 GTS 320 Mb - 5.6 points
Description

Compare that with the futuremark link provided in his post on page 3 showing a 14000+ score, then see that the specs of that rig are not close to the specs here, damn look at the differences in core and memory speed nevermind the 200Mhz on the Quad.

[DarkMatter]

I see a lot of "Since I can't get that high it's impossible to reach there" mentality.
It's because of this that I have to agree with cefurcan, although I don't like his manners (there are better ways to communicate than flaming each other).
He is showing lots of links that prove his point, something that others don't.
In regards to your last link Tatty: as I see it, if at 4 core@4.5Ghz you reach 15000, it's more than feasible to reach 14000 at 3.6Ghz. 8800gts 320 performance is pretty close to that of the 640 version, even closer on low res, low AA/AF, shader intensive scenarios. Also I would say 320 is easier to overclock due to smaller memory densities: same interface+half memory = half density per chip => less heat => more stability).
And final but not least, most people are overlooking the shader power. I don't see any info about the shader clock they are using, wich is the most important thing. In modern games and apps shader power is what drives performance figures and I will never get tired of saying this. Just look at how performs the 8800GT even thogh it has less rops.

[Lekamies]

Originally Posted by Tatty_One

There is no need to start a flame war here between the 2 of you, we are all entitled to our opinions, in this case one thinks that the futuremark score is a single card and legitimate, the other does not, I happen to go with Trt on this one, not because of anything else but the following FACTS that I will post, the supposed futuremark link you posted, you beleive to be a single card setup and had the core clock running at 720Mhz with a shader clock of 1832, on a Q6600 at 3.6Gig, now I do not beleive that a 14000+ score can be acheived on a 320MB GTS single card setup for these reasons........

My 3d mark'06 score @
Q6600@3,6ghz club 3d 8800gts 320MB | gpu:720mhz | shader: 1836mhz | mem: 1080mhz
14094

[DarkMatter]

Originally Posted by Tatty_One

Edit: Taken from HWbot, 14,197, these are the rig specs:

Processor: Core 2 Q6600 (2.4Gh... @ 3800mhz no image
Videocard: GeForce 8800 GTS 32... @ 783/1026mhz
Global Rank: 261st - 5.3 points
Hardware Rank: 7th GeForce 8800 GTS 320 Mb - 5.6 points
Description

Compare that with the futuremark link provided in his post on page 3 showing a 14000+ score, then see that the specs of that rig are not close to the specs here, damn look at the differences in core and memory speed nevermind the 200Mhz on the Quad.

Are you asking how can a lower Core/Memory clocked card win the other?
OK... I will say it AGAIN. SHADER POWER. (Btw Shader Power as Stream Procesor count x Shader Clock x Efficiency).

[cefurkan]

Originally Posted by Lekamies

My 3d mark'06 score @
Q6600@3,6ghz club 3d 8800gts 320MB | gpu:720mhz | shader: 1836mhz | mem: 1080mhz
14094

here is the living expample

so dont bother urself if u cant understand these cards capasity and how much they dependet on cpu mhz also ram timings and ram mhz

[Tatty_One]

OK I am old enuff and bold enuff (although I have plenty of hair) to be the first to stand corrected and eat humble pie But if I may comment on a couple of points raised, firstly.....Darkmatter, I have just skipped through the thread and I cannot find anyone who is saying that they cannot hit 14000 so it cannot be possible, if you look, all my comments were geared towards the 14000 futuremark link, I dont think based on the very threads that you are refering to (but most importantly you didnt look at the one in my post, 5 above this) that showed a HWBot rig Q6600 @ 3.8Gig with a 320MB GTS clocked at a core of 783Mhz (both CPU and core considerably higher than lekamies) that it was unreasonable of me to deduct that at 3.6Gig on a quad with considerably lower clocks that it was not valid??? that rig with those specs only scored 14,1 and something, I know different now but as I said, I think it was a reasonable deduction based on facts.

Yes I understand about Shader clocks, beleive it or not, I have my card shader overclocked as well as core overclocked with the shader not in sync with the core, as you will probably be aware, that is easy now within the latest edition of Rivatuner. My point was, with a core clock of 783Mhz on the rig example I showed, without any further "shader tweaking" the shader clocks are not going to be significantly lower and I would imagine someone who has such a good 3D Mark 2006 score as the example I gave would have enuff knowledge to overclock his shaders seperetly in any case. In any case, we are talking 3D Mark 2006 here which is not particularily shader intensive.

I was never interested in whether the card could acheive scores over 14000, it was quite obvious from the research I did that it was capable (because earlier I did state I could be wrong.....and I was!), I just found it difficult to beleive that those clocks on a Q6600 and the GTS could acheive them, I only really came in because there was a flame war about to start between cefurkan and Trt that was un-necessary, right or wrong, as I said earlier, there is no need to insult people, to be fair he did not insult me but thats not the point.

[trt740]

Originally Posted by Mussels

in my PC a GTX gets 19-19.5K, while a GTS 640 gets around 13.5K

so yes, i beleive your single card score.

Okay prove it bench 3dmarks 06 and show me, no way you can do it I owned a gtx it had the second fastest core clock here and the fastest ram clock. Plus I had a Thermalright h03 air cooler which matches water cooling.Do it then, with a single card and prove it. 6000 point difference no way. Then post it. I do ,however, stand corrected on the 8800 gts it appears it can break 14000+ must have dry Ice strapped to it and a blessing or holy water in it's cooling system, but what do I know. I want to be proven wrong again post that 19000 gtx with the actual screen shots of rivia, cpuz etc and the score. Mussel if you can do then show me. You said your card can do it so lets see it.

[trt740]

Originally Posted by trt740

Here is mine again the top card on this forum so beat post it just like this and your my hero. I have no doubt a 8800 gtx can beat this score but not 19000+. I could care less about futuremarks web site show it to me here and now just like this.

remember this is no average overclock my quad is at 4.041 ghz my card is at 877/2394 and my PICE slot is overclocked from 100mghz to 130 mghz. This is on air.

[trt740]

19000 so prove it. Hes at 3.9ghz so about 50 mghz slower than my cpu

[strick94u]

As my specs show I have a 8800 gts 640 SC oced SLI and decent hardware to back it up and I would have to agree that 19000 is not going to happen even with quad core. so I call bullshi^

[trt740]

Hey it is on future marks it has to be true even without any screenies to back up like the ones I just posted.

[bigboi86]

Originally Posted by AphexDreamer

Indeed.

You're wrong about that though. There is no connection between GPU and CPU speeds.

[trt740]

Originally Posted by strick94u

As my specs show I have a 8800 gts 640 SC oced SLI and decent hardware to back it up and I would have to agree that 19000 is not going to happen even with quad core. so I call bullshi^

[DarkMatter]

3DMark is not shader intensive? Cmon... How can then a less pixel fillrate, less texel fillrate, but much more shader powered HD 2900s be ahead of their respective competitors on this benchmark if they are just behind them on most games and benchies?

As for the other thing I was only trying to point out that on those conditions 14000 were not improbable at all, granted you believe the others at 15000. It's not that a big difference. You even mention on your post that only the CPU clock increase to 4.5Ghz can account for 800 points increase. In the same post you assume (based on core/memory) the shaders are running at 1860Mhz, but who knows? That is what I was trying to explain, I didn't say anywhere you didn't know nothing about shaders, but you oversaw the fact that they could be using a lower shader clock, and automatically you say it's imposible for a single GTS to achieve 14000 at those speeds. I was basically responding to that post on your part.
The "Since I can't reach..." was about other people, not you. That is why I said that first and then I talked to you. Should I have made 2 posts? Maybe. Still there are lots of post saying that they barely reach 14k with their GTX and thus 14k on the GTS is impossible, my first sentence was 4 them.

Sorry for the rant.

[trt740]

Originally Posted by bigboi86

You're wrong about that though. There is no connection between GPU and CPU speeds.

In this bench it does because part of the bench is cpu performance . Regardless your wrong a slow cpu can hold back a gpu it's called bottle necking.

[DarkMatter]

Originally Posted by trt740

In this bench it does because part of the bench is cpu performance . Regardless your wrong a slow cpu can hold back a gpu it's called bottle necking.

There's a big missunderstanding between the camps on this subject. One camp says that you can push both clock frequencies independently, without one keeping the other from going higher.
The other camp says that you can't, when they really want to say that you shouldn't, or that there is a point where past it it doesn't make sense to put one higher without pushing the other at the same time.