• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

X2900XT preview

Whew, this forum reeks......

The HD2900XT should compete with the 8800GTS if we go by any of the past matchings. XT vs. GT/GTS, XTX vs. GTX.

HD2900XT will compete with the 8800GTS boys, plain and simple. If it's close to or better than the GTX that's just a bonus.
 
As I see it both the 8800 series and R600 series are being marketed as DX10 cards. Until there are games out to show which is faster in a DX10 game no one is a winner. Being faster at DX9 games is just an add little bonus. Being first to the market doesnt mean your the best and being unpopular doesnt mean your wrong.

Even if the R600 cards are released on Monday along witha mountain of reviews, those reviews are still just for DX9 games. They wont show who actually has the better DX10 card.
 
debello_64 said:
As I see it both the 8800 series and R600 series are being marketed as DX10 cards. Until there are games out to show which is faster in a DX10 game no one is a winner. Being faster at DX9 games is just an add little bonus. Being first to the market doesnt mean your the best and being unpopular doesnt mean your wrong.

Even if the R600 cards are released on Monday along witha mountain of reviews, those reviews are still just for DX9 games. They wont show who actually has the better DX10 card.
Click to expand...

By the time DX10 games hit in force, we'll be in the second (if not third) generation of DX10 gpus. So right now, DX9 performance is a HUGE matter.
 
and those X3 numbers are BS. EVERY review I have seen the X1950 XTX. Beats, or is a a few FPS (5-8) behind the GTX. We are to believe the 2900 is SLOWER than the X1950 XTX???

hope this is the case.
 
An update to the it "review"; http://it-review.net/index.php?option=com_content&task=view&id=1325&Itemid=91

Still not sold on the authenticity of it however. The M2N32-SLI is a AM2 mobo, yet they show they are using a C2D 6700? They show Oblivion much faster with 8xAA at 2560x1200 than with 4xAA. To the tune of 32fps compared to 51fps. Not to mention many more inconsistencies... as Ive been saying all along, Ill wait for a trusted review.

There's one other "trick" here, and that's image quality, also "to be covered" in a separate article. In all honesty, 8 and 16xAA modes look a bit better on HD2900XT then on 8800 series. This is one of those things that make HD2900XT what it is - very interesting product. Stay tuned for CrossFire vs SLI article tomorrow, done on 975 chipset for CrossFire and 680i chipset for 8800's.
Click to expand...


They do say this about IQ, but again, Ill wait for real reviews...
 
LawGiver said:
and those X3 numbers are BS. EVERY review I have seen the X1950 XTX. Beats, or is a a few FPS (5-8) behind the GTX. We are to believe the 2900 is SLOWER than the X1950 XTX???

hope this is the case.
Click to expand...

wait a second you are saying that the X1950 XTX is faster than the 8800GTX now? :rolleyes:
 
fallguy said:
Still not sold on the authenticity of it however. The M2N32-SLI is a AM2 mobo, yet they show they are using a C2D 6700? They show Oblivion much faster with 8xAA at 2560x1200 than with 4xAA. To the tune of 32fps compared to 51fps. Not to mention many more inconsistencies... as Ive been saying all along, Ill wait for a trusted review.
Click to expand...


Did you miss the bit where it says

So, let's see how R600 does on 975x
Click to expand...

Or

OK, this time, for the 975 test, we picked the ASUS's P5W DH Deluxe motherboard,
Click to expand...
 
FORNOWAGAIN said:
Did you miss the bit where it says

Or
Click to expand...

Nope. Did you miss the part where they said they had a M2N32-SLI. Probably a copy/paste error from another review. But its just one more of many inconsistencies.

Since you want to defend them, explain how they got almost 20 more fps at 8x, than at 4x in Oblivion.
 
Today I found the c't magazine (germanys most credible computer magazine) in my mailbox and they have an article about the 2900XT and some benchmarks. This issue will be in stores on 14th may but as a subscriber I got it earlier :D

Well, what should I say... The card is a big disappointment. The whole article is like a verification for all the bad news and benchmarks that we have seen in the last days :(

Only in three tests which are based on the DX10 SDK that card is unbelievable fast.
 
Shoggy said:
Today I found the c't magazine (germanys most credible computer magazine) in my mailbox and they have an article about the 2900XT and some benchmarks. This issue will be in stores on 14th may but as a subscriber I got it earlier :D

Well, what should I say... The card is a big disappointment. The whole article is like a verification for all the bad news and benchmarks that we have seen in the last days :(

Only in three tests which are based on the DX10 SDK that card is unbelievable fast.
Click to expand...

If you could please post a picture so that we finally have FACTS that prove what has been reported all along and then the fanboys can figure out some excuse for this article being fake too...

edit:
And if it is true ATi made a big mistake going for DX10 performance this round (and not attempting to get good DX 9 performance) since by the time games are out for DX10 NVidia will have released G90 most likely..
 
Shoggy said:
Today I found the c't magazine (germanys most credible computer magazine) in my mailbox and they have an article about the 2900XT and some benchmarks. This issue will be in stores on 14th may but as a subscriber I got it earlier :D

Well, what should I say... The card is a big disappointment. The whole article is like a verification for all the bad news and benchmarks that we have seen in the last days :(

Only in three tests which are based on the DX10 SDK that card is unbelievable fast.
Click to expand...

http://www.xtremesystems.org/forums/showpost.php?p=2185477&postcount=912

Wombat is pretty reputable. I dont doubt his word on it.
 
Shoggy said:
Today I found the c't magazine (germanys most credible computer magazine) in my mailbox and they have an article about the 2900XT and some benchmarks. This issue will be in stores on 14th may but as a subscriber I got it earlier :D

Well, what should I say... The card is a big disappointment. The whole article is like a verification for all the bad news and benchmarks that we have seen in the last days :(

Only in three tests which are based on the DX10 SDK that card is unbelievable fast.
Click to expand...

What kind of disappointment? Disappointing that it only competes with the GTS (as it should)? Or disappointing that it can't compete with the GTS?
 
JustinSane said:
What kind of disappointment? Disappointing that it only competes with the GTS (as it should)? Or disappointing that it can't compete with the GTS?
Click to expand...

Hahaha, as it should? Are you kidding me? Have you seen the specs of the XT? It seems ATI's $400 pricing is having the desired effect :)
 
Shoggy said:
Have you read his first sentence *g* -> "...from europs biggest PC-Magazine ct".
That's the article I'm talking about ;)
Click to expand...

I know, I'm just saying if people wanted scans, theres another person whose seen the article. :)

Two people confirming it should be more than enough for its existence.
 
JustinSane said:
GTS320 = 2900
GTS640 = XT
GTX = XTX
Click to expand...

No idea what you're trying to say here....the XT and XTX have typically been within 5% of each other. The GTX is 40-50% faster than the GTS. You trying to say that ATI planned to clock the XTX at over 1100Mhz on 80nm? Because that's what it would have to be clocked to in order to line up with the GTX.

Honestly, anybody with a little technical knowledge and some common sense would realize that the XT is a full R600 that simply isn't performing up to its potential. Hence AMD had no choice but to cut the price. It was never aimed at GTS performance, that's a joke.
 
To add some perspective, taking the GTS as a baseline -

MADD flops
XT: +108%
GTX: +50%

Bandwidth
XT: +65%
GTX: + 32%

Texturing (assuming the reported 16 bilinear TMU's for r600 is correct)
XT: +0%
GTX: +53%

So what picture does that paint? That ATI designed a monster chip with great specs. And then hobbled it with lackluster texturing performance. And that last bit is just a guess - it's texturing performance could very well be higher depending on the structure of its TMU's.

So if you still think ATI was aiming at GTS performance with 65% more bandwidth and 2x the MADD flops you might want to re-think that opinion. I'm hoping better drivers reveal the true performance of this part. In which case $400 will be a real deal.
 
Shoggy said:
Today I found the c't magazine (germanys most credible computer magazine) in my mailbox and they have an article about the 2900XT and some benchmarks. This issue will be in stores on 14th may but as a subscriber I got it earlier :D
Click to expand...
Keep in mind that there is a significant delay between authoring and publishing for print magazines.
Well, what should I say... The card is a big disappointment. The whole article is like a verification for all the bad news and benchmarks that we have seen in the last days :(
Click to expand...
That it is marginally faster than the same price range nVidia card, the GTS640?
Only in three tests which are based on the DX10 SDK that card is unbelievable fast.
Click to expand...
Curious. What kind of test was that? A 3DMark kind of test or some demo sequence or such?
 
LawGiver said:
Shoggy :

POST PICTURES OF THE MAGAZINE WE WANT TO SEE TOO YOU BASTARD!!!!:D
Click to expand...

Sure, if you pay me when their huge law department is sending a "nice" letter... :p

Dwight, it's only an article with two sites and the last sentence says that more details will follow in the next issue. I'm pretty sure they have done the benchmarks short before the deadline (wouldn't be the first time) - and that was definitely within this week. That mag comes out every 14 days. So the gap can't be sooooo huge...

We have seen a lot of leaked benchmarks before and now we have some first official benchmarks (a few days old) which can only confirm all the rumors. So do you really think on monday the card will be any better?!

And btw: I bet that all the sites which will have a review on monday, have already done their tests. So their benchmarks are "old" too.
 
We have seen a lot of leaked benchmarks before and now we have some first official benchmarks (a few days old) which can only confirm all the rumors.
Click to expand...
Given the perpensity of the rumours to contradict each other, even if you narrow it down to the "bad" ones.....umm, yeah. ;)
 
Dwight said:
That it is marginally faster than the same price range nVidia card, the GTS640?
Click to expand...

No that it is marginally faster than a much weaker card hardware wise. And obviously people weren't waiting all this time for performance that they could have had months ago. Why wouldn't they be disappointed?
 
trinibwoy said:
No that it is marginally faster than a much weaker card hardware wise.
Click to expand...
Umm, OK. Exactly how did you come up with "weaker"? Oh yeah, maybe by comparing some numbers....in what is an apparently erroneous manner. Otherwise how is this much weaker card not? :eek:

P.S. A "weaker" card with 128MB more memory?
 
....arm-chair interpreted in an apparently erroneous way. *shrug* It is like saying "Hey, this 30cm ruler isn't any longer than that 12 inch one. But 30>>12! WHAT A FIASCO!!!!!" :rolleyes:
 
Shoggy said:
Have you read his first sentence *g* -> "...from europs biggest PC-Magazine ct".
That's the article I'm talking about ;)
Click to expand...

Wow... just wow. The performance numbers in that magazine article are pathetic for the R600 :(.
 
Dwight said:
in what is an apparently erroneous manner.
Click to expand...

Heh, please correct me then. I'm awaiting your great insight into GPU hardware to set me right :) Of course, I won't hold my breath. Because if you had the slightest understanding of how the hardware works you would understand my point.

The fact that you think the GTS is faster because of 128MB more memory is proof enough of the level of your understanding though. Which is to say, not very high :p
 
trinibwoy said:
Heh, please correct me then. I'm awaiting your great insight into GPU hardware to set me right :)
Click to expand...
*shrug* Don't need to. There it is! Learn.
Of course, I won't hold my breath.
Click to expand...
I wouldn't expect you to since that would involve you stopping your gum flapping. :cool:
Because if you had the slightest understanding of how the hardware works you would understand my point.
Click to expand...
:rolleyes:
The fact that you think the GTS is faster because of 128MB more memory is proof enough of the level of your understanding though. Which is to say, not very high :p
Click to expand...
You missed my point. The memory given the card can provide an insight into where it was aimed in terms of applications.

P.S. Oh, and yes extra RAM can greatly increase performance under some applications....didn't yah know? :confused:
 
fallguy said:
Nope. Did you miss the part where they said they had a M2N32-SLI. Probably a copy/paste error from another review. But its just one more of many inconsistencies.

Since you want to defend them, explain how they got almost 20 more fps at 8x, than at 4x in Oblivion.
Click to expand...
Don't be obtuse. Pointing out, as badly as its written, an express statement that the x975 is used is hardly defending them.
 
The last few weeks have been quite the roller coaster ride for those of us looking forward to the release of the new cards. Things should be much clearer on Monday.
Personally, I'm much more interested in what the cards performance is in DX10. However, there only appear to be a few apps available for benching. Of those the XT seems to be doing quite well.
 
Dwight said:
P.S. Oh, and yes extra RAM can greatly increase performance under some applications....didn't yah know? :confused:
Click to expand...


as I said in a previous post (wasn't pretaining to any of your posts) only when page flipping is occuring.
 
It's really strange that this card isn't really performing much better than an x1950xt. I'll be interested to see what driver updates can do for this card. Considering how different the GPU is from the x1900xt, it seems like something is missing from these benchmarks. It's just puzzling when you have a card with more than double the stream processors, double the memory bandwidth, and yet the performance just isn't there. DX10, maybe?

Regardless though, I don't think that people were expecting a card that competes with the 8800GTS, or even one that is close to an 8800GTX. I think people were expecting a card that would really trounce the 8800GTX and take the performance crown from graphzilla. The GPU is clearly quite well-designed but something seems to be holding it back. So, any way you slice it, it's a disappointment. There's always the possibility of competitive pricing, but not only are prices unconfirmed, but for all we know nVidia could cut prices on their cards when the r600 finally hits shelves.
 
FORNOWAGAIN said:
Don't be obtuse. Pointing out, as badly as its written, an express statement that the x975 is used is hardly defending them.
Click to expand...

Take your own advice. The simple fact is, they claimed to have used an AMD board, with an Intel chip. Which is one of the inconsistencies they have.
 
You must log in or register to reply here.
Back
Top