GT 280/260 performance comparison to 3870X2

[Pinacolada]

LMAO ! Nvidia guy who did this first chart should go to primary school to learn some math, what a misleading marketing.Shame on you,nvidia.

[[XC] hipno650]

The 4870 is not twice as fast as the 3870 so the 4870X2 will definitely not be twice as fast as a 3870X2.

In fact, early drivers for the 4870X2 (which is July we're talking about btw) will most likely have minor scaling issues just as the 3870X2 did, it will have a smaller performance delta than the 4870 vs the 3870.

The 4870 AT MOST (with huge amounts of wishful thinking) will only be about 50% faster than the 3870 in games and real world applications. It will be a lot faster in benchmarks but I don't "play" benchies,

finally another has seen the light!!! like i have said before the 4870 is not going to be to much faster than the 9800gtx and that will put the 4870x2 a little faster than the 9800gx2 so now it's AMD's new gen competing with nvidia's "old gen". but one thing i for see in both cards is new drivers will make a large performance increase as time passes on.

[r1ch]

LMAO ! Nvidia guy who did this first chart should go to primary school to learn some math, what a misleading marketing.Shame on you,nvidia.

"Misleading marketing" is a bit of misnoma, no?

Marketing is intended to be misleading, why would nVidia produce a slide that showed a 'fair' comparison with a smaller gap? Would more people buy it because they're being underwhelmingly honest? No.

I'm not defending their flagrantly biased use of scale, lack of numbers, or test scenario detail to back up the data, but this is exactly how the marketing departments of BOTH companies work. Unfortunately there are n00bs who will believe this graph.

Likewise, there are some who will believe the 'normalized' graph that someone produced in the HD4000/GTX280 threads that plotted a 65% increase across the board for the 4870 card. That, was equally as misleading.

Suck it all up and come to your own conclusions. Trust those websites who will be open and honest with the settings, scale and provide proof.

[Eastcoasthandle]

"Misleading marketing" is a bit of misnoma, no?

Marketing is intended to be misleading, why would nVidia produce a slide that showed a 'fair' comparison with a smaller gap? Would more people buy it because they're being underwhelmingly honest? No.

I'm not defending their flagrantly biased use of scale, lack of numbers, or test scenario detail to back up the data, but this is exactly how the marketing departments of BOTH companies work. Unfortunately there are n00bs who will believe this graph.

Likewise, there are some who will believe the 'normalized' graph that someone produced in the HD4000/GTX280 threads that plotted a 65% increase across the board for the 4870 card. That, was equally as misleading.

Suck it all up and come to your own conclusions. Trust those websites who will be open and honest with the settings, scale and provide proof.

This is not correct. Marketing primary goal is to bring public awareness to what is available for sale. Not to mislead the public

[Papu]

that would be propaganda ?

[bill_d]

nv marketing primary goal is to make the GTX Titanic look good

[Shintai]

nv marketing primary goal is to make the GTX Titanic look good

They dont need marketing for that.

[r1ch]

This is not correct. Marketing primary goal is to bring public awareness to what is available for sale. Not to mislead the public

Exactly. If they'd produced the graph with a true scale then it wouldn't have caught so many people's eyes. We wouldn't be talking about it, and it wouldn't have it's own thread. In that respect, job done for nVidia.

Tbh, the only people who are being mislead are those who are dumb enough not to realise the scale (not to mention trusting a biased source).

[Isaac MM]

Do you people think that CrossFire-X HD 4870X2 will beat a GTX 280?

[Luka_Aveiro]

Do you people think that CrossFire-X HD 4870X2 will beat a GTX 280?

Even a single HD4870X2 could beat a GTX280 (or not), just wait for the truth instead of feeding yourself with fantasy

[informal]

Do you people think that CrossFire-X HD 4870X2 will beat a GTX 280?

There is a definite possibility that it can beat it(from game to game since cards/GPUs act differently over the range of game engines).
IF 4870 is indeed 40% faster than 9800GTX,then 4870X2 will be around 40%(or some less) faster than 9800GX2 which is in line with GTX280 performance expectations(around 30% faster than 9800GX2).

[bill_d]

can someone just press the fastward button till august

[generics_user]

can someone just press the fastward button till august

QFT

[ColonelCain]

I am still waiting for a true, up-front comparison before I make any descisions, or claims for that matter. All that I know, is that 2008/2009 is going to be really fun!

[Calmatory]

3x GTX 280 vs. 2x HD4870X2.

[Mr.Guvernment]

...though, the benchmarks are done internally by Nvidia....

nuff said...... BS, will wait for some real numbers.

[Xello]

3x GTX 280 vs. 2x HD4870X2.

That's what I'm talkin' about.

[Gunslinger]

already posted in the 9900GTX & GTS thread in sticky - http://www.xtremesystems.org/forums/...185628&page=35

First, thanks for fixing that badly skewed chart. Typical marketing mumbo jumbo.
As for posting in the other thread, the new cards are not named 9900GTX & GTS, so maybe that thread should die off.

[Boissez]

Before people here starts ooh-ing and aah-ing over those scores they should look at the little green logo in the top right corner and consider the following:

1. Would Nvidia include games that favor ATI in those slides?
2. Would Nvidia use settings that favor ATI cards?
3. Would Nvidia use drivers that favor ATI cards?
4. Would Nvidia present these data in a favorable way to ATI?

If you've answered yes to any of those questions I urge you to have another long look at that little green logo and consider the source. While I do not doubt that the GTX's will kill the 3870X2 these data cannot/must not/should not be taken at face value.

FWIW the 3870X2 might actually beat the GTX 280 in WIC as there is a 2,6X speed-up in going from Cat 8.3 to 8.4:
http://images.tweaktown.com/imagebank/cat84_g_11.gif
[source]

[Decami]

this looks like an fps chart for GFX cards from 1983, does anyone know what in the hell this chart means or even represents besides the "we are better than them" theory. Someone needs to let Nvidia know that only Albert Einstein and Isaac Newton are allowed to show us charts that we dont understand!

[Luka_Aveiro]

Before people here starts ooh-ing and aah-ing over those scores they should look at the little green logo in the top right corner and consider the following:

1. Would Nvidia include games that favor ATI in those slides?
2. Would Nvidia use settings that favor ATI cards?
3. Would Nvidia use drivers that favor ATI cards?
4. Would Nvidia present these data in a favorable way to ATI?

If you've answered yes to any of those questions I urge you to have another long look at that little green logo and consider the source. While I do not doubt that the GTX's will kill the 3870X2 these data cannot/must not/should not be taken at face value.

FWIW the 3870X2 might actually beat the GTX 280 in WIC as there is a 2,6X speed-up in going from Cat 8.3 to 8.4:
http://images.tweaktown.com/imagebank/cat84_g_11.gif
[source]

This ruined a perfectly sarcastic-unbiased post

[Mr.Guvernment]

nuff said...... BS, will wait for some real numbers.

Before people here starts ooh-ing and aah-ing over those scores they should look at the little green logo in the top right corner and consider the following:

1. Would Nvidia include games that favor ATI in those slides?
2. Would Nvidia use settings that favor ATI cards?
3. Would Nvidia use drivers that favor ATI cards?
4. Would Nvidia present these data in a favorable way to ATI?

If you've answered yes to any of those questions I urge you to have another long look at that little green logo and consider the source. While I do not doubt that the GTX's will kill the 3870X2 these data cannot/must not/should not be taken at face value.

FWIW the 3870X2 might actually beat the GTX 280 in WIC as there is a 2,6X speed-up in going from Cat 8.3 to 8.4:
http://images.tweaktown.com/imagebank/cat84_g_11.gif
[source]

exactly..... logo = charts means nothing

[gOJDO]

1) I agree with many posters here. Shame on nVidia, they should give a credit to AMD for the chart style
2) We have no info about the hardware, the OS(es) and the software used. Also we have no info about the resolutions used in that comparison. So we can safely assume that the chart is representing the best case for the nVidia cards and/or the worst case for the 3870X2.
3) nVidia used outdated and unoptimized ATi drivers, while they haven't specified the drivers for their new cards. nVidia was cheating in the past and we don't know if they are not going to do the same again with the drivers for their new cards.

The chart only gives a simple image about the performance of GT200 in the best case. This month we'll see a lot of benches done by the lucky individuals who have connections with nVidia.

[r1ch]

3x GTX 280 vs. 2x HD4870X2.

In performance, 3x GTX 280.
In price/performance, probably 2x HD 4870X2.

[annihilat0r]

Why should 4870 be only %50 (and even that "with a lot of wishful thinking") faster than 3870 when its processing power doubled, memory bandwidth doubled, memory amount doubled, texture addressing/filtering more than doubled, and z-fillrate at least doubled?

And if 4870 is %50 better than 3870, which makes 4850 only %10 or 15 better than 3870; how the hell can AMD/ATi justify 4850's price of $220, whereas HD3870's are as low as $150?

I would expect 4870 to be close to twice 3870, which would justify both part's prices and specifications.