HD 2900XTX Benches

[HurleyBird]

Too many emoticons, apoppin. Like some preteen on MSN...

 

[chizow]

Originally posted by: ShadowOfMyself
Wow... You people even know Kombatant? He tells the truth, no more, no less, Ill take his word over DT any day

Nope I don't know him, not that it matters. What exactly was he saying over the last 6-12 months that was so truthful?

'We got this really kick @ss card guys...its well worth the wait....all you ATI fans are gonna be so happy......you are just not ready for it yet.....'

k. Not his quotes, but I'm going to assume he didn't say ANYTHING, much less anything that could show he always told the truth, the whole truth, and nothing but the truth over the last 6 months about R600. Sounds like he's just towing the company line and playing the part of damage control at this point.

I just see ATI trying to put out hundreds of fires as a result of these DT results. No official word on anything for months and once these results leak there's all these "credible" sources coming out of the woodwork saying DT's results are off. Its AMD's own fault for putting themselves in this position. If they had a target product and apprised the press about it from the start half of these rumors wouldn't exist.

Here's a short list of some of the AMD fumbles so far:

1) Delays because of a unified product launch. Totally BS excuse with the XT finally coming to soft-launch with the mid-range and XTX limping along on a TBA launch date.

2) Dragonhead and whatever code-names for the 2900 variants. AFAIK, the 1GB GDDR4 was ALWAYS slated to be the 2900XTX and the part DT got, whether ATI wants to say its an OEM XT or not, has 1GB GDDR4. But who knows since ATI never officially said anything to lock them into this position.

3) Basing XTX expectations on target clockspeeds and XT oc'ing headroom. Pointless because the XTX has been pushed back for who knows how long and the DT XT OC'ing results showed almost no increase whatsoever in terms of actual performance. This may change with more thorough testing, but right now, basing XTX performance on potential OCs is pointless.

I could go on, but you get the point. This R600 situation is ridiculous and its only getting worst lol.

 

[HurleyBird]

Originally posted by: swtethan
BTW, gtx's run 650/2000 oc on AIR easily


mine runs 630/2100 on stock AIR

70c load

The thing is that overclocking to those levels and releasing at those levels is completely different. In order to release at those levels you need to be stable even in bad conditions. Thats why you see the 8800 Ultra with its new monster heat sink. According to you Nvidia should be able to release the 8800 Ultra with the old GTX PCB and heat sink.... riiiiight.

 

[Matt2]

Originally posted by: ShadowOfMyself
Wow... You people even know Kombatant? He tells the truth, no more, no less, Ill take his word over DT any day

Looks like everyones just enjoying jumping at the AMD hate bandwagon, sort of like Sony

Well, i am glad i listened to Kyle

Dont ever do that again

You would rather believe an employee of AMD about it's performance over a third party?

Hell, if Rollo were here he'd probably tell you 8600GTS was a kick ass card and you guys would tell him to climb a tree.

But this Kombatant guy says something and we're supposed to take his word over DailyTechs?

 

[chizow]

Originally posted by: HurleyBird

Originally posted by: swtethan
BTW, gtx's run 650/2000 oc on AIR easily


mine runs 630/2100 on stock AIR

70c load

The thing is that overclocking to those levels and releasing at those levels is completely different. In order to release at those levels you need to be stable even in bad conditions. Thats why you see the 8800 Ultra with its new monster heat sink. According to you Nvidia should be able to release the 8800 Ultra with the old GTX PCB and heat sink.... riiiiight.

But you do realize that's exactly what AMD and its supporters are trying to do with the XTX in the latest wave of DT rebuttals right? I do agree it was unfair to bench the OC'd GTX to the OEM XTX (fake one, according to our reliable AMD sources lol) if the XTX is going to release with higher clockspeeds, but who's fault is it for not releasing actual specs? Personally I'll go with the hardware provided now, not AMD's word on what they planned or intended on the "real" XTX.

 

[HurleyBird]

Originally posted by: Matt2

Originally posted by: ShadowOfMyself
Wow... You people even know Kombatant? He tells the truth, no more, no less, Ill take his word over DT any day

Looks like everyones just enjoying jumping at the AMD hate bandwagon, sort of like Sony

Well, i am glad i listened to Kyle

Dont ever do that again

You would rather believe an employee of AMD about it's performance over a third party?

Hell, if Rollo were here he'd probably tell you 8600GTS was a kick ass card and you guys would tell him to climb a tree.

But this Kombatant guy says something and we're supposed to take his word over DailyTechs?

Seeing as other people under NDAs have called the numbers bunk and DT made rank amateur mistakes in both reviews?

I'd say neither.

On one hand, you can't take the ATI employee seriously because he's an ATI employee. On the other hand, you can't take the DT numbers as authoritative because of aforementioned problems with the 'reviews'.

The only move that you can take that proves you aren't an idiot fanboy is to wait until the NDA expires.

 

[HurleyBird]

Originally posted by: chizow
But you do realize that's exactly what AMD and its supporters are trying to do with the XTX in the latest wave of DT rebuttals right?

Oh, I realize that. It's a bit of a different situation though. The XT overclocks by 100MHz on a weaker cooler, so it's pretty rational to claim 50MHz on a stronger cooler as reasonable. Claiming that the XTX will launch at 850MHz is silly though (even though it may be possible depending on differences between tested silicon and production silicon, if any).

 

[chizow]

Originally posted by: HurleyBird

Originally posted by: chizow
But you do realize that's exactly what AMD and its supporters are trying to do with the XTX in the latest wave of DT rebuttals right?

Oh, I realize that. It's a bit of a different situation though. The XT overclocks by 100MHz on a weaker cooler, so it's pretty rational to claim 50MHz on a stronger cooler as reasonable. Claiming that the XTX will launch at 850MHz is silly though (even though it may be possible depending on differences between tested silicon and production silicon, if any).

Swtethan didn't claim anything different actually. He's saying its easy to achieve a 650MHz OC on a GTX (only 75MHz iirc) with a stock cooler. You're saying they can't release it at that speed because they need to build in a safety buffer for worst-case scenario situations. The reviewed OC GTX was at 650MHz with I'm assuming a better cooler, and I wouldn't go as far to say you could get another 50MHz out of it, just like I wouldn't assume you could get another 50MHz out of the XTX with a better cooler. In both cases, the better cooler might just be enough to get the cards running stably at those clockspeeds eating up any potential headroom in the process.

 

[HurleyBird]

Originally posted by: chizow

Originally posted by: HurleyBird

Originally posted by: chizow
But you do realize that's exactly what AMD and its supporters are trying to do with the XTX in the latest wave of DT rebuttals right?

Oh, I realize that. It's a bit of a different situation though. The XT overclocks by 100MHz on a weaker cooler, so it's pretty rational to claim 50MHz on a stronger cooler as reasonable. Claiming that the XTX will launch at 850MHz is silly though (even though it may be possible depending on differences between tested silicon and production silicon, if any).

Swtethan didn't claim anything different actually.

The way he said it showed that he thought that OC ability = release ability.

 

[chizow]

Originally posted by: HurleyBird
The way he said it showed that he thought that OC ability = release ability.

Well, I'm not going to spend too much time on this since it comes down to a matter of interpretation I suppose. But in the end, you're assuming the XTX + better cooler has additional OC'ing headroom where the GTX + better cooler does not. Given the R600's delays, reported heat, yield and power consumption issues, and my own past overclocking experiences, I'm not going to make that assumption. I guess we can follow up on this one in 3-4 months when the XTX releases, if it does at all.

 

[swtethan]

Originally posted by: HurleyBird

Originally posted by: chizow

Originally posted by: HurleyBird

Originally posted by: chizow
But you do realize that's exactly what AMD and its supporters are trying to do with the XTX in the latest wave of DT rebuttals right?

Oh, I realize that. It's a bit of a different situation though. The XT overclocks by 100MHz on a weaker cooler, so it's pretty rational to claim 50MHz on a stronger cooler as reasonable. Claiming that the XTX will launch at 850MHz is silly though (even though it may be possible depending on differences between tested silicon and production silicon, if any).

Swtethan didn't claim anything different actually.

The way he said it showed that he thought that OC ability = release ability.

factory OC = release

 

[HurleyBird]

Originally posted by: chizow

Originally posted by: HurleyBird
The way he said it showed that he thought that OC ability = release ability.

Well, I'm not going to spend too much time on this since it comes down to a matter of interpretation I suppose. But in the end, you're assuming the XTX + better cooler has additional OC'ing headroom where the GTX + better cooler does not.

Nope, both cases should give better ocing headroom. I never said otherwise.


[/quote]
factory OC = release[/quote]

What? That's like saying cheerios = release, heh. Completely unrelated.

In any case, your argument was that you could OC the GTX to 650/1000 easily. My argument is that OC ability does not equal release ability. The only GTX to clock that high (other than the unreleased one DT seems to have) is water cooled. Just because you can sometimes OC that high on the stock cooler doesn't mean you can release at that level on the stock cooler. You need to account for unideal conditions.

 

[aka1nas]

Originally posted by: HurleyBird

Originally posted by: chizow

Originally posted by: HurleyBird
The way he said it showed that he thought that OC ability = release ability.

Well, I'm not going to spend too much time on this since it comes down to a matter of interpretation I suppose. But in the end, you're assuming the XTX + better cooler has additional OC'ing headroom where the GTX + better cooler does not.

Nope, both cases should give better ocing headroom. I never said otherwise.

factory OC = release[/quote]

What? That's like saying cheerios = release, heh. Completely unrelated.

In any case, your argument was that you could OC the GTX to 650/1000 easily. My argument is that OC ability does not equal release ability. The only GTX to clock that high (other than the unreleased one DT seems to have) is water cooled. Just because you can sometimes OC that high on the stock cooler doesn't mean you can release at that level on the stock cooler. You need to account for unideal conditions.[/quote]

More importantly, the only GTX that you can purchase at that clock speed costs around $900. I wouldn't really have had an issue if they had either compared the XT/XTX with a standard GTX OC (~620Mhz core), or had made it more clear what they were comparing in the benchmarks.

 

[HurleyBird]

Originally posted by: aka1nas

Originally posted by: HurleyBird

Originally posted by: chizow

Originally posted by: HurleyBird
The way he said it showed that he thought that OC ability = release ability.

Well, I'm not going to spend too much time on this since it comes down to a matter of interpretation I suppose. But in the end, you're assuming the XTX + better cooler has additional OC'ing headroom where the GTX + better cooler does not.

Nope, both cases should give better ocing headroom. I never said otherwise.

factory OC = release

What? That's like saying cheerios = release, heh. Completely unrelated.

In any case, your argument was that you could OC the GTX to 650/1000 easily. My argument is that OC ability does not equal release ability. The only GTX to clock that high (other than the unreleased one DT seems to have) is water cooled. Just because you can sometimes OC that high on the stock cooler doesn't mean you can release at that level on the stock cooler. You need to account for unideal conditions.[/quote]

More importantly, the only GTX that you can purchase at that clock speed costs around $900.[/quote]

And that doesn't even include the water cooler, just the block.

 

[aka1nas]

Hehe, I have a friend who has tendency to ask me to spec out a new rig for him, and then goes out and buys something that has at least double those specs. The last time, he ended up with a QX6700, 4GB of RAM, and 2 of those watercooled BFG GTXs in SLI. I think I recommended a single 8800GTS.

 

[chizow]

Originally posted by: HurleyBird

Originally posted by: chizow

Originally posted by: HurleyBird
The way he said it showed that he thought that OC ability = release ability.

Well, I'm not going to spend too much time on this since it comes down to a matter of interpretation I suppose. But in the end, you're assuming the XTX + better cooler has additional OC'ing headroom where the GTX + better cooler does not.

Nope, both cases should give better ocing headroom. I never said otherwise.

factory OC = release[/quote]

What? That's like saying cheerios = release, heh. Completely unrelated.

In any case, your argument was that you could OC the GTX to 650/1000 easily. My argument is that OC ability does not equal release ability. The only GTX to clock that high (other than the unreleased one DT seems to have) is water cooled. Just because you can sometimes OC that high on the stock cooler doesn't mean you can release at that level on the stock cooler. You need to account for unideal conditions.[/quote]

Here we go again....apples to apples and oranges to oranges please. If he can't assume the GTX can release with an OC @ 650 based on stock cooler results, then you can't make the same assumption with the XTX based on the XT OC'ing results, no matter what the actual OC'd versions are using for cooling. The XT reviews clearly show there's only so much you can do with cooling at this point, since they're nearly identical to the 8800 coolers.

Judging from your comments you're insinuating the card reviewed by DT was water cooled and costs $900? lol....I don't think many are going to take that leap considering the shortcoming of the reviews along with the early results being limited to 1280x1024 because "they didn't have a monitor that could go higher." There's a few that get close to 650MHz on more extravagant air coolers, but I'm not sure if DT OC'd them more or the card is yet unreleased.

 

[ShadowOfMyself]

More speculation from B3D

From the cryptic messages from Kombatant, and looking at comments about Crossfire vs GTX, 24xAA, improved crossfire solution for R600, better image quality on the R600, no XTX model will be released, you will be suprised by the performance compared to 8800, etc, I can only come to the following conclusion:

AMD will launch the following to compete against 8800 GTS and 8800 GTX:

~$399 HD 2900 XT 512MB RAM
~$699 CROSSFIRE HD 2900 XT 1028MB RAM (basically two the above in crossfire, at $50 discount each)

 

[chizow]

Originally posted by: ShadowOfMyself
More speculation from B3D

From the cryptic messages from Kombatant, and looking at comments about Crossfire vs GTX, 24xAA, improved crossfire solution for R600, better image quality on the R600, no XTX model will be released, you will be suprised by the performance compared to 8800, etc, I can only come to the following conclusion:

AMD will launch the following to compete against 8800 GTS and 8800 GTX:

~$399 HD 2900 XT 512MB RAM
~$699 CROSSFIRE HD 2900 XT 1028MB RAM (basically two the above in crossfire, at $50 discount each)

LOL! Why is CF even being discussed without mention of SLi? Seriously, anyone else find this really pathetic that they're trying to compare 2 XTs in CF vs. 1 GTX without explicitly saying so? Is this the big surprise that Kombatant is talking about? That 2 XT in CF perform better than 1 GTX? That guy is not only truthful, he's a genius. They list the price, but they'res no mention its a 2 v 1 scenario. Might as well list 2 GTS vs those 2 XT if you're looking for a price: performance ratio with multiple cards.

 

[aka1nas]

Originally posted by: chizow

Judging from your comments you're insinuating the card reviewed by DT was water cooled and costs $900? lol....I don't think many are going to take that leap considering the shortcoming of the reviews along with the early results being limited to 1280x1024 because "they didn't have a monitor that could go higher." There's a few that get close to 650MHz on more extravagant air coolers, but I'm not sure if DT OC'd them more or the card is yet unreleased.

No, he's stating that they compared the XTX to a card whose specs currently can only be bought by end users at the $900 price point. rather than the $550 GTX at regular clock speeds or the slightly more expensive OC'ed GTXs that run around 620Mhz on the core. The fact that some or many GTXs can hit that speed when OC'ed on air is irrelevent as they weren't attempting to test how far each card can OC nor did they run the XTX at higher clocks.

 

[chizow]

Originally posted by: aka1nas
The fact that some or many GTXs can hit that speed when OC'ed on air is irrelevent as they weren't attempting to test how far each card can OC nor did they run the XTX at higher clocks.

Re-read the argument. That's all thats relevant because he said you can't assume a GTX OC on air with a stock cooler can sufficiently satisfy safe working conditions to release at that speed. Then ofc the ATI camp goes on to assume the XTX will obviously release at higher clock speeds based on the XT OC'ing results with a stock cooler....

I don't disagree that a OC vs. OC comparison would've been more appropriate as I've said numerous times, but again, based on the XT OC'ing results I doubt it would've made much difference.

 

[ShadowOfMyself]

Originally posted by: chizow

Originally posted by: ShadowOfMyself
More speculation from B3D

From the cryptic messages from Kombatant, and looking at comments about Crossfire vs GTX, 24xAA, improved crossfire solution for R600, better image quality on the R600, no XTX model will be released, you will be suprised by the performance compared to 8800, etc, I can only come to the following conclusion:

AMD will launch the following to compete against 8800 GTS and 8800 GTX:

~$399 HD 2900 XT 512MB RAM
~$699 CROSSFIRE HD 2900 XT 1028MB RAM (basically two the above in crossfire, at $50 discount each)

LOL! Why is CF even being discussed without mention of SLi? Seriously, anyone else find this really pathetic that they're trying to compare 2 XTs in CF vs. 1 GTX without explicitly saying so? Is this the big surprise that Kombatant is talking about? That 2 XT in CF perform better than 1 GTX? That guy is not only truthful, he's a genius. They list the price, but they'res no mention its a 2 v 1 scenario. Might as well list 2 GTS vs those 2 XT if you're looking for a price: performance ratio with multiple cards.

And Nvidia was the first to do it... Since the 7900 wasnt up to the task of defeating the 1950, they launched the 7950GX2 as if it were one card

Back then I thought it was lame, so if AMD does the same, I still think its lame

 

[chizow]

Originally posted by: ShadowOfMyself

Originally posted by: chizow

Originally posted by: ShadowOfMyself
More speculation from B3D

From the cryptic messages from Kombatant, and looking at comments about Crossfire vs GTX, 24xAA, improved crossfire solution for R600, better image quality on the R600, no XTX model will be released, you will be suprised by the performance compared to 8800, etc, I can only come to the following conclusion:

AMD will launch the following to compete against 8800 GTS and 8800 GTX:

~$399 HD 2900 XT 512MB RAM
~$699 CROSSFIRE HD 2900 XT 1028MB RAM (basically two the above in crossfire, at $50 discount each)

LOL! Why is CF even being discussed without mention of SLi? Seriously, anyone else find this really pathetic that they're trying to compare 2 XTs in CF vs. 1 GTX without explicitly saying so? Is this the big surprise that Kombatant is talking about? That 2 XT in CF perform better than 1 GTX? That guy is not only truthful, he's a genius. They list the price, but they'res no mention its a 2 v 1 scenario. Might as well list 2 GTS vs those 2 XT if you're looking for a price: performance ratio with multiple cards.

And Nvidia was the first to do it... Since the 7900 wasnt up to the task of defeating the 1950, they launched the 7950GX2 as if it were one card

Back then I thought it was lame, so if AMD does the same, I still think its lame

Actually the situation was different, the 7900 was released before the 1950, the GX2 was only NV's attempt at retaking the crowd. And yes, SLi/CF cards being compared to single card solution is lame, but most here can make the distinction between a single and duo configuration.

In this scenario, the XT in CF seems to be AMD's answer to the GTX without addressing 2 GTX in SLi since it seems it will be a 2 card solution and not a GX2-type, effectively limiting the configuration to 2 GPU vs. 2 GPU and not 4 GPU vs. 2 GPU.

Anyways, gonna go save Middle Earth for a while and wait for better benchmarks on Tuesday. Have a nice weekend all.

 

[HurleyBird]

Originally posted by: chizow
Here we go again....apples to apples and oranges to oranges please. If he can't assume the GTX can release with an OC @ 650 based on stock cooler results, then you can't make the same assumption with the XTX based on the XT OC'ing results, no matter what the actual OC'd versions are using for cooling. The XT reviews clearly show there's only so much you can do with cooling at this point, since they're nearly identical to the 8800 coolers.

Judging from your comments you're insinuating the card reviewed by DT was water cooled and costs $900? lol....I don't think many are going to take that leap considering the shortcoming of the reviews along with the early results being limited to 1280x1024 because "they didn't have a monitor that could go higher." There's a few that get close to 650MHz on more extravagant air coolers, but I'm not sure if DT OC'd them more or the card is yet unreleased.

Learn to read before you criticize, wise guy.

I never claimed that ATI could release an XTX at the same clocks as DT was able to achieve on the XT. I did state that half of that amount should feasible though. In fact, so did DT.

As for insinuating that DT was using a water cooled card, it so happens that the *only* 8800 GTX on the market that sports the same clocks as the card DT used is water cooled. However, a DT employee told me that the card they used wasn't water cooled. In other words, it's likely that DT is using an unreleased air cooled 8800 GTX with a huge factory overclock. I already posted this info earlier in the thread. Go back and read it.

Originally posted by: chizow
Re-read the argument. That's all thats relevant because he said you can't assume a GTX OC on air with a stock cooler can sufficiently satisfy safe working conditions to release at that speed.

I never claimed that. I've only pointed out that the only card released at those speeds is water cooled. Obviously a bigger air cooler can also do the trick. Just look at the 8800 Ultra.

 

[Munky]

What retaking the crown with the 7950gx2? Prior to that the 7900gtx never had the crown in the first place, even against the x1900xtx. The fact that it took two g71 chips to beat a single r580 wasn't too impressive, and if Ati does the same with the r600, then it basically confirms that a single r600 can't match a single g80.

Also, about the whole OC thing, 650mhz is not exactly easy for the 8800gtx. Looking at some oc results here, here, here, here, here, and here, only a minority of the tested cards reached 650 and higher.

 

[chizow]

Originally posted by: HurleyBird

Originally posted by: swtethan
BTW, gtx's run 650/2000 oc on AIR easily


mine runs 630/2100 on stock AIR

70c load

The thing is that overclocking to those levels and releasing at those levels is completely different. In order to release at those levels you need to be stable even in bad conditions. Thats why you see the 8800 Ultra with its new monster heat sink. According to you Nvidia should be able to release the 8800 Ultra with the old GTX PCB and heat sink.... riiiiight.

Learn to read....k. That's your original post on this topic (I didn't bother responding to your other ad hominem attacks elsewhere), to which I replied that is exactly what people are basing higher clocked XTX expectations on. You replied:

Oh, I realize that. It's a bit of a different situation though. The XT overclocks by 100MHz on a weaker cooler, so it's pretty rational to claim 50MHz on a stronger cooler as reasonable. Claiming that the XTX will launch at 850MHz is silly though (even though it may be possible depending on differences between tested silicon and production silicon, if any).

The GTX can't hit higher clock speeds on a stock cooler because of safe working limits but we can expect the XTX to? Apples to apples.....