HD 2900XTX Benches

Page 8 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

razor2025

Diamond Member
May 24, 2002
3,010
0
71
Originally posted by: n7
Originally posted by: n7

I suspect the poor performance is due to the stream processors layout for the AMD cards vs. nV's, but i can't find anything stating just how it's set up.

From a poster commenting on the DT article:

This is what i was thinking too.

This is very interesting bit of information and shines light on the reason why AMD bought ATI in the first place. If anyone's following CPU/GPU/GPxPU news, you can see that things are starting to converge upon each other. From the programmable shaders in GF3, GPU has started to walk the path into becoming a "general purpose" processor, albeit a bit more specialized than CPU. Folding with ATI's GPU already proved that GPUs are quite capable of running non-visual code. If indeed what that post says is true, then I understand AMD's strategy for next few years, especially with the "Torrenza" and "HT 3.0 Co-processors". Think super integration/linking. As much as current situation looks bleak for AMD, if their bet goes well with this strategy, they can edge out Intel in whole different way(why do you think Intel's been hyping out their in-house GPU so much lately?).

Anyways, I would wait until official numbers or 2-3 more sources giving the same numbers as DT do, before I pass judgement down on R600. I'm not even interested in cards past $300 segment, as very few will buy them + they're not the profit for GPU makers. Sub $200 and especially sub $100 market is where the money is made for GPUs. If you doubt me, you can easily check Valve's hardware surveys. Add to the fact that 86xx series are disappointing with their price segment. If the HD 26xx series outperforms 86xx series maintaining the same price range, than that's all it matters for me.
 

Nightmare225

Golden Member
May 20, 2006
1,661
0
0
Damn, this would really suck for ATI.

On the other hand, this means that my $600 purchase was an amazing choice and that my 8800GTX is officially the longest card in the history of the 3D gaming market to hold the performance crown.
 

Extelleron

Diamond Member
Dec 26, 2005
3,127
0
71
Originally posted by: razor2025
Originally posted by: n7
Originally posted by: n7

I suspect the poor performance is due to the stream processors layout for the AMD cards vs. nV's, but i can't find anything stating just how it's set up.

From a poster commenting on the DT article:

This is what i was thinking too.

This is very interesting bit of information and shines light on the reason why AMD bought ATI in the first place. If anyone's following CPU/GPU/GPxPU news, you can see that things are starting to converge upon each other. From the programmable shaders in GF3, GPU has started to walk the path into becoming a "general purpose" processor, albeit a bit more specialized than CPU. Folding with ATI's GPU already proved that GPUs are quite capable of running non-visual code. If indeed what that post says is true, then I understand AMD's strategy for next few years, especially with the "Torrenza" and "HT 3.0 Co-processors". Think super integration/linking. As much as current situation looks bleak for AMD, if their bet goes well with this strategy, they can edge out Intel in whole different way(why do you think Intel's been hyping out their in-house GPU so much lately?).

Anyways, I would wait until official numbers or 2-3 more sources giving the same numbers as DT do, before I pass judgement down on R600. I'm not even interested in cards past $300 segment, as very few will buy them + they're not the profit for GPU makers. Sub $200 and especially sub $100 market is where the money is made for GPUs. If you doubt me, you can easily check Valve's hardware surveys. Add to the fact that 86xx series are disappointing with their price segment. If the HD 26xx series outperforms 86xx series maintaining the same price range, than that's all it matters for me.

Regardless of the fact that most can't afford them, people do look to the ultra-high end for their purchasing decisions. Every gamer used to want a 7800GTX and now its the 8800GTX... when they can't afford that they will go with the same company's mid-range or low-end GPU. So, in a way, high-end performance can influence sales of lower-end hardware.

 

Ackmed

Diamond Member
Oct 1, 2003
8,478
524
126
Originally posted by: PingSpike
I'm really hoping Creig is right about this and there's something weird going on, because this is really the last thing AMD needs right now.

Here is some no-name site claiming to have a Crysis demo, and the X2900XT. beating the 8800GTX. I wouldnt take these numbers with any sort of seriousness though.

I dont know if there is anything funny or not going on. I was hoping it would be equal, or a little faster than the 8800GTX to be honest. Enough to keep both companies trying to out do one another the next round. I will wait for real reviews before I pass judgment however. These one page, few game, no IQ comparison, no talk about new feature reviews just dont do it for me. If they are in-line with others that will be out, its not good for people who waited. Unless its way cheaper, and has some features the 8800 series do not have.
 

SsupernovaE

Golden Member
Dec 12, 2006
1,128
0
76
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.

You know, it's probably that those rendering programs will scale almost perfectly with a vector-based engine. Games, on the other hand, seem to be a lot more finicky. If there is indeed more efficiency in the DX 10 API, then it's conceivable that the R600 could stomp all over the G80, at least theoretically.
 

GTaudiophile

Lifer
Oct 24, 2000
29,776
31
81
So it DT trying to steal some theInq's thunder by spreading FUD...using its Anandtech parent as some sort of validity to their claims?
 

Nightmare225

Golden Member
May 20, 2006
1,661
0
0
Originally posted by: GTaudiophile
So it DT trying to steal some theInq's thunder by spreading FUD...using its Anandtech parent as some sort of validity to their claims?

Unlike theInq, they have actual "benchmarks".
 

Drayvn

Golden Member
Jun 23, 2004
1,008
0
0
Originally posted by: Creig
The more I think about it, the more I'm convinced that something just isn't right with numbers DT is providing. Saying that ATI is releasing a new card that can't keep up with an 8800GTX is one thing. Saying that ATI is releasing a new card that is barely faster than their previous generation champ (X1950XTX), even when using a quad core processor, is another.

Company of heroes 1280*1024:
1950 XTX 99 (tweaktown) E6600 (dual core)
2900 XTX 97 (dailytech) QX6800 (quad core)

Company of heroes 1600*1200:
1950 XTX 70 (tweaktown) E6600 (dual core)
2900 XTX 73 (dailytech) QX6800 (quad core)


FEAR 1280*1024:
1950 XTX 80 (tomshardware) No softshadow 4AA 8AF , X6800 (dual core)
2900 XTX 84 (dailytech) with softshadow 4AA 16AF , QX6800 (quad core)

FEAR 1600*1200:
1950 XTX 57 (tomshardware) No softshadow 4AA 8AF , X6800 (dual core)
2900 XTX 58 (dailytech) with softshadow 4AA 16AF , QX6800 (quad core)


Gibbo over at OcUK had the following to say after benching a 2900XT and comparing it to his 8800GTX OC:

Hi there

I cannot say too much, but the card is well built, good quality, is quiet even when underload and is cooler running then previous ATI cards.

Performance wise everyone will have to wait but I think everyone will be impressed with its performance, especially if suggested pricing from other threads and articles from other websites are to be believed. I really cannot say anything or I will get my ass kicked from AMD. As soon as I can you guys will know more.


Something


just


doesn't


add


up.

Exactly,

And if you look at the AT benches of the XFX 8800GTX XXX with the 630Mhz Core 2000Mhz Memory. In Oblivion its getting HALF the FPS that DT benched their slightly higher OCed card. I dont 20Mhz can do that much difference.

Also the Half Life 2 FPS are completely wrong, as for some reason that highly OCed card is getting about HALF the FPS of the AT reviewed one!

As you said

Something

Just

Doesn't

Add

Up

P.S. Also didn't they do this with the X1800 or X1900 when it was late arriving or something and they got the numbers completely wrong?
 

dreddfunk

Senior member
Jun 30, 2005
358
0
0
Chest thumping and hand wringing...a flagship launch must be upon us...I love the smell of flaring egos in the morning...err...afternoon.

Why don't we all just treat predicting these things as an intellectual exercise rather than a statement of our own personal value and call it a day? It's fascinating to ponder where things are headed; there's no need to take it personally when someone doesn't agree with you as long as they aren't offensive about doing so (which, admittedly, is usually the problem).

On to the topic at hand.

If the different shader model actually gives R600 an advantage in DX10 titles, I'm not sure it was the worst call ever made, but the way the timing worked out does make it look awfully bad right now. That's a big 'if', however. I won't be convinced of any of this nonsense until there is a lot of data--both DX9 and DX10--on the subject.

At any rate, the design process for these chips takes quite a while. It's very conceivable that when ATI made the decision to go with their shader setup that they thought DX10 titles would coincide more closely with Vista's launch. By the time it was evident that they wouldn't, they may have been stuck with a seemingly bad design decision.

Again, we'll wait for the real DX10 numbers, but if there is a big jump in the HD2900's performance, it may have been just a risky gamble by ATI that didn't pay off. If DX10 titles were currently available, there would probably be a decent number of gamers that would trade a little DX9 performance for some extra DX10 performance.

Still...as it stands, this is a potential train wreck in the making. I don't see this design decision helping me in the midrange--even if the 2600 has great DX10 performance because of the different shader setup, doesn't seem likely to have any better DX9 performance than the 8600, which means that my aging x850xt is likely to remain in service a while longer.

At least until there is a strong sub-$200 nVidia part that can tackle Oblivion with HDR and 4xAA.
 

Genx87

Lifer
Apr 8, 2002
41,095
513
126
Originally posted by: SsupernovaE
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.

You know, it's probably that those rendering programs will scale almost perfectly with a vector-based engine. Games, on the other hand, seem to be a lot more finicky. If there is indeed more efficiency in the DX 10 API, then it's conceivable that the R600 could stomp all over the G80, at least theoretically.


Well that and it beat the 5500FX, not the 5600FX. The 5500FX is a G70 based card.


 

manowar821

Diamond Member
Mar 1, 2007
6,063
0
0
Originally posted by: dreddfunk
Chest thumping and hand wringing...a flagship launch must be upon us...I love the smell of flaring egos in the morning...err...afternoon.

Damn did I LOL :laugh:
 

flexy

Diamond Member
Sep 28, 2001
8,464
155
106
those cards get owned by the GTX....if thats even remotely the case and no FUAD, ATI doesnt even need to bothero to release that card.

I THOUGHT R600 would be a killer card.

Acording to those numbers GTX eats a R600 for breakfast.

Glad i dont have stocks in ATI....VERY DISAPPOINTED
 

Extelleron

Diamond Member
Dec 26, 2005
3,127
0
71
Originally posted by: flexy
those cards get owned by the GTX....if thats even remotely the case and no FUAD, ATI doesnt even need to bothero to release that card.

I THOUGHT R600 would be a killer card.

Acording to those numbers GTX eats a R600 for breakfast.

Glad i dont have stocks in ATI....VERY DISAPPOINTED

If those "tests" are correct the R600 is a few percentage points faster than an X1950XTX. Since that's definately not true, you can tell the whole thing is BS. The HD 2900XT/X will not crush the 8800GTX but it will not be slower, either, and definately not by this much.
 

swtethan

Diamond Member
Aug 5, 2005
9,083
0
0
Originally posted by: Extelleron
Originally posted by: flexy
those cards get owned by the GTX....if thats even remotely the case and no FUAD, ATI doesnt even need to bothero to release that card.

I THOUGHT R600 would be a killer card.

Acording to those numbers GTX eats a R600 for breakfast.

Glad i dont have stocks in ATI....VERY DISAPPOINTED

If those "tests" are correct the R600 is a few percentage points faster than an X1950XTX. Since that's definately not true, you can tell the whole thing is BS. The HD 2900XT/X will not crush the 8800GTX but it will not be slower, either, and definately not by this much.

oh so youre an ATI engineer are you? :lol;
 

secretanchitman

Diamond Member
Apr 11, 2001
9,352
21
91
its looking like this is real, but i dunno. im kinda skeptical, but not really...

my 8800GTS 320MB doesnt look so bad anymore (like it ever was).
 

Extelleron

Diamond Member
Dec 26, 2005
3,127
0
71
Originally posted by: swtethan
Originally posted by: Extelleron
Originally posted by: flexy
those cards get owned by the GTX....if thats even remotely the case and no FUAD, ATI doesnt even need to bothero to release that card.

I THOUGHT R600 would be a killer card.

Acording to those numbers GTX eats a R600 for breakfast.

Glad i dont have stocks in ATI....VERY DISAPPOINTED

If those "tests" are correct the R600 is a few percentage points faster than an X1950XTX. Since that's definately not true, you can tell the whole thing is BS. The HD 2900XT/X will not crush the 8800GTX but it will not be slower, either, and definately not by this much.

oh so youre an ATI engineer are you? :lol;

No, I'm a person who has looked at these numbers and seen how crazy they are. At most the 8800GTX is 2x as fast as the X1950XTX. In some of those tests, the 8800GTX is not much less than 2x as fast as the R600XTX. Does that not sound fishy to you?
 

swtethan

Diamond Member
Aug 5, 2005
9,083
0
0
Originally posted by: Extelleron
Originally posted by: swtethan
Originally posted by: Extelleron
Originally posted by: flexy
those cards get owned by the GTX....if thats even remotely the case and no FUAD, ATI doesnt even need to bothero to release that card.

I THOUGHT R600 would be a killer card.

Acording to those numbers GTX eats a R600 for breakfast.

Glad i dont have stocks in ATI....VERY DISAPPOINTED

If those "tests" are correct the R600 is a few percentage points faster than an X1950XTX. Since that's definately not true, you can tell the whole thing is BS. The HD 2900XT/X will not crush the 8800GTX but it will not be slower, either, and definately not by this much.

oh so youre an ATI engineer are you? :lol;

No, I'm a person who has looked at these numbers and seen how crazy they are. At most the 8800GTX is 2x as fast as the X1950XTX. In some of those tests, the 8800GTX is not much less than 2x as fast as the R600XTX. Does that not sound fishy to you?

did they test a 1950xtx on the same system? same timedemo?

plus, since the new drivers came out, how much more of an improvement?



 

flexy

Diamond Member
Sep 28, 2001
8,464
155
106
"problem" is the source of those news is reliable

So...what has more weight:

1) "Wishful thinking" (That this is only FUD because it CANT be what CANT be
2) Just accept the truth and SEE the reason why they had that LONG delay - because R600 is a flop,
and see those benchmark numbers from a RELIABLE (!) source, namely AT/DT.

Sorry kids, 2) is MORE likely...things add up. Everything else is just wishful thinking.

Next

 

Extelleron

Diamond Member
Dec 26, 2005
3,127
0
71
Originally posted by: flexy
"problem" is the source of those news is reliable

So...what has more weight:

1) "Wishful thinking" (That this is only FUD because it CANT be what CANT be
2) Just accept the truth and SEE the reason why they had that LONG delay - because R600 is a flop,
and see those benchmark numbers from a RELIABLE (!) source, namely AT/DT.

Sorry kids, 2) is MORE likely...things add up. Everything else is just wishful thinking.

Next


I don't care how much of a flop the R600 is, a company does not design a new GPU architecture to be 10-20% faster than their old flagship. ATI has worked on R600 for years, and has released a similar chip with R500. They have experience with a DX10 (well, R500 isn't really DX10 but its close) architecture and have had years of development time. There is no way that the top-end HD 2900XTX is going to be only slightly faster than an X1950XTX.
 

Genx87

Lifer
Apr 8, 2002
41,095
513
126
Originally posted by: Extelleron
Originally posted by: flexy
those cards get owned by the GTX....if thats even remotely the case and no FUAD, ATI doesnt even need to bothero to release that card.

I THOUGHT R600 would be a killer card.

Acording to those numbers GTX eats a R600 for breakfast.

Glad i dont have stocks in ATI....VERY DISAPPOINTED

If those "tests" are correct the R600 is a few percentage points faster than an X1950XTX. Since that's definately not true, you can tell the whole thing is BS. The HD 2900XT/X will not crush the 8800GTX but it will not be slower, either, and definately not by this much.


We will have to wait and see of course. But this post sounded a lot like the Nvidia fanboys(including myself) right before the NV30 launch
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |