HD 2900XTX Benches

[tuteja1986]

Originally posted by: otispunkmeyer

Originally posted by: 5150Joker

Originally posted by: chrismr

Originally posted by: tuteja1986
Something is seriously wrong with the benchmark because if its right than a Dual X1950pro is faster than X2900XTX in those benchmark.

And if thats the case, then maybe the 1950 dual doesn't look like such a bad idea anymore - if the price is right.

I'd rather get the X2900XT since it's at least DX10.

and if the XT can do some good video decoding and encoding then its still a good choice.
hell if it encodes DivX thats more than a reason for me to get one

Well according to ATI X2900XT can beat up a FX5600 $3000 workstation card in CAD , Maya...ect.

 

[otispunkmeyer]

Originally posted by: rmed64

Originally posted by: otispunkmeyer

Originally posted by: HigherGround

Oh and, yay for testing stock 2900 agaisnt overclocked 8800 :roll:

why is everybody bitchin' and moaning about that. This is exactly what happens when you get late to a party, you end up going against the fastest RETAIL card available, not a REFERENCE design.

not only that they didnt have time to go find a stock card.

what would you rather them do? produce 1 or 2 tests while they fanny'd around looking for a stock card or downlcocking the one they had on a machine which might not even be theirs or do you wan them to get cracking on some numbers?

seems to me DT are damned if they do and damned if they dont.

people would complain of lack of benches, or they would complain of rushed benches....sorry but given the time frame they had, they did the best they could.

still its not that big an overclock and some ones already pointed out that even at stock it still had no hope of beating the 8800GTX (little bit of extrapolation tho)

and higher ground is right.

ok it might be fair to do stock to stock, but the fact of the matter is, when you go to buy a 8800GTX you cant help but trip over countless vendor tuned models....infact i know shops where thats the only choice you have..... BFG OC's. EVGA superclocked KO's, golden samples etc etc. whats the point in benchin it against a stock 8800 when thats not what people are buying?

what they should be testing really is popular cards, or testing at the same price point. for me its all about bang for buck. if Nvidia have the faster card at the time simply because EVGA overclocked it for me...thats the one ill buy, im not interested that stock vs stock it isnt the fastest because im not gonna go home and then turn the wick down on cool bits am i?

this is just simply the price ATi must pay for being late. OC'd editions are the cards of choice for 8800 buyers...thats just a fact, ATi will only have refrence clocked cards to sell..and if past experience is anything to go by thats all they'll have becuase i dont think ATi have ever let vendors jimmy the clocks much if at all.

If these benches are true, there is no point in releasing the XTX, just release the XT for $399 and be done with it.

When testing, to be fair, if they are gonna use an oced GTX, they should of oced the XTX to the same percentage (if possible).

well no they shouldnt because its not representative of whats actually on offer to buy. if i can get a 8800GTX OC for £400 or a HD2900XTX for £400...then i should be looking at how a OC'd 8800 compares with a stock HD2900. not stock vs stock.

if all you can buy in the shops is a reference XT or a OC'd GTX then thats how they should be tested. its pretty much useless to any buyer benching something that doesnt exist.

same if it was the other way round, if nvidias card were all sold at stock clocks, but ATi let their vendors OC to their hearts content, then it would have to be stock nvidia vs OC ATI

if you wanna test fairly to see how one cores architecture is better than another...then really it shouldnt be stock vs stock.....it should clock for clock.... both at the same speeds.

 

[otispunkmeyer]

Originally posted by: tuteja1986

Originally posted by: otispunkmeyer

Originally posted by: 5150Joker

Originally posted by: chrismr

Originally posted by: tuteja1986
Something is seriously wrong with the benchmark because if its right than a Dual X1950pro is faster than X2900XTX in those benchmark.

And if thats the case, then maybe the 1950 dual doesn't look like such a bad idea anymore - if the price is right.

I'd rather get the X2900XT since it's at least DX10.

and if the XT can do some good video decoding and encoding then its still a good choice.
hell if it encodes DivX thats more than a reason for me to get one

Well according to ATI X2900XT can beat up a FX5600 $3000 workstation card in CAD , Maya...ect.

whats that got to do with Divx/xvid/H.264 encoding?

 

[aka1nas]

I wish they would give better details on what they benched at. They were really lazy with the Oblivion benchmarks in particular. For all we know, the X2900s were running 24x AA and the GTX was running 8x. I'm also getting seriously pissed that they won't bench with HDR+AA. Why would I spend $400+ on any of those cards to just run one or the other?

The XT is still looking like it will be a great card if it comes out to be somewhere around $400. I wasn't even anywhere near in the market for an XTX, so I'm still hopeful.

 

[5150Joker]

I'm still curious about how ATi's new cards fare in DX10 games and tests. DX9 is old hat and that's all we've really seen of both the 8800 series and now the R600, who knows what will happen in DX10.

 

[Ulfhednar]

It's starting to look like I will be keeping my BFG 8800GTS 640MB until 65nm GPUs come out.

Or, at the very least, until Crysis comes out and Vista gets some driver and patch support.

 

[aka1nas]

Originally posted by: otispunkmeyer
well no they shouldnt because its not representative of whats actually on offer to buy. if i can get a 8800GTX OC for £400 or a HD2900XTX for £400...then i should be looking at how a OC'd 8800 compares with a stock HD2900. not stock vs stock.

if all you can buy in the shops is a reference XT or a OC'd GTX then thats how they should be tested. its pretty much useless to any buyer benching something that doesnt exist.

same if it was the other way round, if nvidias card were all sold at stock clocks, but ATi let their vendors OC to their hearts content, then it would have to be stock nvidia vs OC ATI

if you wanna test fairly to see how one cores architecture is better than another...then really it shouldnt be stock vs stock.....it should clock for clock.... both at the same speeds.

The only problem with that is the only 650Mhz GTX that you can actually buy is a water-cooled BFG model that costs over $900. It's somewhat silly to present that as the typical GTX that someone would actually purchase.

 

[Jules]

Originally posted by: aka1nas

Originally posted by: otispunkmeyer
well no they shouldnt because its not representative of whats actually on offer to buy. if i can get a 8800GTX OC for £400 or a HD2900XTX for £400...then i should be looking at how a OC'd 8800 compares with a stock HD2900. not stock vs stock.

if all you can buy in the shops is a reference XT or a OC'd GTX then thats how they should be tested. its pretty much useless to any buyer benching something that doesnt exist.

same if it was the other way round, if nvidias card were all sold at stock clocks, but ATi let their vendors OC to their hearts content, then it would have to be stock nvidia vs OC ATI

if you wanna test fairly to see how one cores architecture is better than another...then really it shouldnt be stock vs stock.....it should clock for clock.... both at the same speeds.

The only problem with that is the only 650Mhz GTX that you can actually buy is a water-cooled BFG model that costs over $900. It's somewhat silly to present that as the typical GTX that someone would actually purchase.

Or a BFG 8800GTX thats 570$ AR that does 626MHz

 

[RichUK]

Is the size of the HD 2900 XTX supposed to be some kind of joke?

 

[jim1976]

Meh.. I call shens.. INQ,FUDzilla and Dailytech..

 

[sisq0kidd]

Originally posted by: RichUK
Is the size of the HD 2900 XTX supposed to be some kind of joke?

Well, it does say it's mainly for OEM machines, which would already take into consideration the size.

Originally posted by: jim1976
Meh.. I call shens.. INQ,FUDzilla and Dailytech..

I don't believe Dailytech is lying at all or being disingenuous with their benchmarks in the slightest.

They benched what they received. It's a driver issue on ATI's part... I hope

 

[RichUK]

Originally posted by: sisq0kidd

Originally posted by: RichUK
Is the size of the HD 2900 XTX supposed to be some kind of joke?

Well, it does say it's mainly for OEM machines, which would already take into consideration the size.

It makes no difference whether it's an OEM card or not. That's still the reference size of the XTX, and it F'ing huge.

Unless I?m missing something here and the retail cards will actually be smaller?

 

[BlizzardOne]

Retail will be 9.5" like the XT.

The OEM is that big because that's how big OEM's want it. They just made the cooler bigger to satisfy OEM requirements (see the OEM 7800GTX with the 3" handle on the end).

 

[RichUK]

Originally posted by: BlizzardOne
Retail will be 9.5" like the XT.

The OEM is that big because that's how big OEM's want it. They just made the cooler bigger to satisfy OEM requirements (see the OEM 7800GTX with the 3" handle on the end).

That would explain it, thanks.

 

[Rommel44]

FAKE

 

[jim1976]

Originally posted by: sisq0kidd
I don't believe Dailytech is lying at all or being disingenuous with their benchmarks in the slightest.

They benched what they received. It's a driver issue on ATI's part... I hope

I used to respect them but not anymore..But that's just me I suppose..
Also I find weird that in 3dmark scores the X2900XT was dominating and in games falls so much behind.. No matter how much I don't trust 3DMock as an indication of real gaming performance it just doesn't add up.. Nm I'll wait for some more previews/reviews b4 I make my opinion about X2900..

 

[jim1976]

To fire things up even more FUDzilla goodness

http://www.fudzilla.com/index.php?option=com_content&task=view&id=703&Itemid=1

ATI said in front of 150+ journalists that R600XT won't be able to compete with Geforce 8800 GTX. The dream is dead and Radeon HD 2900 XT won't be able to catch up with the six month old Geforce 8800 GTX.

ATI said that R600XT, Radeon HD 2900XT has about the same performance as the 8800 GTS. Radeon HD 2900 XTX won't be launched at mid may and will be delayed to Q3 2007. We already wrote about this here.

The dream that AMD can do something and return as the high end market leader is gone. R600 rails against G80. Congratulations Nvidia.

 

[TanisHalfElven]

Originally posted by: josh6079
God, 2007 has really sucked for us gamers so far.

nope it sucks for amd. 8800gts320 ftw

 

[bunnyfubbles]

Originally posted by: Matt2

Originally posted by: ShadowOfMyself
Hum.... Very disappointing but I want to see AA and HDR tested, everyone knows Nvidia has always had more "Raw" power, and Ati more "efficient" AA performance

Oh and, yay for testing stock 2900 agaisnt overclocked 8800 :roll:

Before anyone asks, the quality settings for the games were as follows:

Company of Heroes - High shader quality, High model quality, Anti-aliasing enabled (in game), Ultra texture quality, high quality shadows, high quality reflections, Post processing On, High building detail, High physics, high tree quality, High terrain detail, Ultra effects fidelity, Ultra effects density, Object scarring enabled and the model detail slider all the way to the right.

F.E.A.R. - 4x FSAA (in game), maximum light details, shadows enabled, maximum shadow details, soft shadows enabled, 16x anisotropic filtering, maximum texture resolution, maximum videos, maximum shader quality.

Half Life 2: Episode 1 - High model detail, high texture detail, high shader detail, reflect all water details, high shadow detail, 4x multi-sample AA (in-game), 16x anisotropic filtering, v-sync disabled, full high-dynamic range.

Elder Scrolls IV Oblivion ? Preset ultra quality settings. The scene used was where you?re at the end of the tunnel and run outside for a little bit.

I'm sorry, but turning up the AA isnt going to make up for 20 fps deficits.

Actually it would (or should), 4x AA isn't exactly anything too aggressive by today's standards. By increasing things that should really put a strain on memory bandwidth, the 2900s should take less and less of a hit while the 8800s would take more and more (relative to each other)

I'm pretty wary of these results as well, the performance hit going to higher resolutions on the 8800GTX is less than that of the 2900s which goes against what we'd expect considering the clear memory bandwidth advantage the 2900s have. What's more is that the XTX's even greater memory bandwidth advantage seems to offer next to nothing when compared to the XT. I smell something fishy.

I'm just hoping this is user error and/or a case of horrendously crappy early drivers like ATI had with the 8500. The GPU might not be what we expected, and I can accept that, but I was really hoping more out of the memory bandwidth.

 

[EKTELESTES]

His benchmarks are a piece of crap, check this thread.

http://forums.guru3d.com/showthread.php?t=222721

And look at the silly results compared to yesterday's XT vs GTS 640 review.

http://www.imageshack.gr/viewer.php?file=tindlzdmomjh0dzm1mxi.jpg]

[/URL]


Check Oblivion results and FEAR and have a good laugh.

 

[GTaudiophile]

Well, AMD/ATI is dead. LOL! Congrats to them!

 

[tuteja1986]

Dude its like you guys are saying X2900XT is dead when it will be only for sold for $399 which is way cheaper than than 8800GTX which goes for $550.

 

[Modular]

Is anyone really surprised that this card is a dud? All signs have been pointing to a failure on launch of this card for quite a long time now: no pre-released benches, NDA, saying they have the thing ready to ship "today" but not doing so, delayed release dates (again by the way), etc, etc, etc. At least a few strongly allied persons will be buying these cards, but it's actually a disadvantage to do so.

Sure, sure, we should all wait for more benches to conclude the same thing as the DT ones before we make any rash decisions, right? It can't be that Ati really dropped the ball. This is a sad day for competition, but not an unexpected one. Maybe nVidia will play nice and lower prices anyway...just wishful thinking I suppose.

Oh and to the poster saying that DX9 is "Old Hat" I completely disagree. Name me one working DX10 title that they could benchmark on, just one. DX9 is far from Old Hat..

 

[HigherGround]

Originally posted by: jim1976
To fire things up even more FUDzilla goodness

http://www.fudzilla.com/index.php?option=com_content&task=view&id=703&Itemid=1

ATI said in front of 150+ journalists that R600XT won't be able to compete with Geforce 8800 GTX. The dream is dead and Radeon HD 2900 XT won't be able to catch up with the six month old Geforce 8800 GTX.

ATI said that R600XT, Radeon HD 2900XT has about the same performance as the 8800 GTS. Radeon HD 2900 XTX won't be launched at mid may and will be delayed to Q3 2007. We already wrote about this here.

The dream that AMD can do something and return as the high end market leader is gone. R600 rails against G80. Congratulations Nvidia.

now we know why they decided to launch it in a middle of a desert

 

[ShadowOfMyself]

This just cant be... I mean, WTF? Even the 1950XTX sometimes catches or comes close to the GTS 640... This has to be a joke