More information about R9 290X

[chimaxi83]

Whatever the hour there s a morning somewhere...

I know, but, for example, the 7970/280X launch reviews came out at 12AM EST, whereas the GTX680/780/Titan reviews came out at 9AM EST. I just noticed those times, so maybe we might read something tonight

 

[Abwx]

I know, but, for example, the 7970/280X launch reviews came out at 12AM EST, whereas the GTX680/780/Titan reviews came out at 9AM EST. I just noticed those times, so maybe we might read something tonight

Two hours or so left and we ll know...

They d better deliver at least thoses reviews in the waiting
of actual manufacturers products.

 

[lavaheadache]

In theory, if the NDA was bumped back could people in the know choose to adhere to the original NDA that they signed and ignore the delay? What could AMD do? I would imagine nothing.

I am aware that everything I said could be wrong since those slides could have been fake to begin with.

 

[AnandThenMan]

I'm sure one of the terms of the NDA are that the NDA can be modified at any time and you must adhere to the updated terms.

 

[lavaheadache]

I'm sure one of the terms of the NDA are that the NDA can be modified at any time and you must adhere to the updated terms.

lol, gotcha.. :thumbsdown:

 

[Cloudfire777]

R9 290X in 3DMark 11 Extreme preset. Notice the Graphic score.

FireStrike Extreme. Also notice Graphic score.

GPU-Z

Many other pictures here:
http://videocardz.com/46785/amd-radeon-r9-290x-performance-charts-emerge

 

[lavaheadache]

I'll just chalk this one up to the ever growing list of dissapointments associated with this launch


*EDIT*

Just to clarify, my comment was regarding the OCT 15th NDA lift.

 

[Grooveriding]

hmmmm

 

[Tweak155]

In theory, if the NDA was bumped back could people in the know choose to adhere to the original NDA that they signed and ignore the delay? What could AMD do? I would imagine nothing.

I am aware that everything I said could be wrong since those slides could have been fake to begin with.

Not send them the card next time

 

[Cloudfire777]

^That is the 290X overclocked if you haven`t noticed.
Yes, GPU-Z is not playing nice with the new not fully supported GPU.

That temp is a dead give away

 

[Tweak155]

I don't ever use these programs... which scores are we looking at? I think it is obvious for the Fire Strike because there are only 2 scores, but what about 3DMark11? Top or bottom?

 

[Cloudfire777]

Its this one that is stock clocks.

R9 290X is about equal with Titan in 3DMark FireStrike Extreme but is about 10% behind on 3DMark11 Extreme. Not sure how that will translate in real gaming performance..

 

[Borealis7]

i don't read chinese...is the graphics score the 1st or the 3rd number?

 

[Grooveriding]

I don't ever use these programs... which scores are we looking at? I think it is obvious for the Fire Strike because there are only 2 scores, but what about 3DMark11? Top or bottom?

I tend to agree. The synthetics are way too random, too heavily driver optimized and often don't represent game performance properly. They are a good way to gauge how cards line up against other cards from the same architecture though; ie. AMD GCN or nvidia Kepler against the same.

The more interesting possible true leaks are the ones of performance in games.

I was hoping the release was today, but since AMD always seems to lift NDAs at 12:00AM EST, I guess it is next Tuesday ?

 

[Borealis7]

i think one of the GPU-Z pics is fake. notice the inconsistencies: shader number, die size, memory size, bandwidth etc'.
it's the same version of gpu-z!

 

[Cloudfire777]

i don't read chinese...is the graphics score the 1st or the 3rd number?

Graphic score is the number below the total score.

3DMark11 Extreme graphic score: 4192
3DMark FireStrike Extreme graphic: 4976

Its the most important score because its basically only GPU performance translated in to points. The total score is not accurate, because if you use a faster CPU than the other GPUs that was tested, the total score will go up. Graphic score will not.

 

[Abwx]

That temp is a dead give away

Fortunately (or unfortunately.??:sneaky GPUZ/CPUZ are known
to give random values for temps and are not reliable
at all , not even for voltages , better find another
tool for thoses parameters...

 

[Cloudfire777]

http://www.anandtech.com/show/7424/amd-radeon-r9-290x-performance-preview-bioshock-infinite

http://www.tomshardware.com/news/amd-radeon-r9-290x-benchmark-performance,24732.html

 

[Triglet]

17 Games Compared (PCTuning)

 

[Silverforce11]

17 Games Compared (PCTuning)

Thats ridiculous, it won every game out of 17.. some of which are known to be better on NV GPUs as well.

24% faster than a 780 is nuts.

 

[Gigglinpig]

Damn. The GTX 780 feels like a bad purchase now. Time to sell the 780 acx, and buy a 290x once the custom factory O.C cards come out.

I went from a 7970 to GTX 780. I have no brand loyalty. Back to AMD for now.

 

[96Firebird]

17 Games Compared (PCTuning)

From the other thread...

These are AMDs' numbers not pctunings' so I won't be calling it just yet

 

[Cloudfire777]

From the other thread...

There is 10% difference between Titan and 780 in 2560x1600p. Since R9 290X is said to offer Titan performance, I have trouble believing 290X beating Titan by additional 16%.

Very impressive if true though.

EDIT: PCTuning number for Bioshock and Tomb Raider is pretty much identical with Anandtech`s/legitreviews review. Could actually be legit

 

[96Firebird]

The fact that they are only showing results at 4K resolution concerns me as to its performance at 1080/1200/1440/1600. Obviously they are showing their strength versus the competitor, but it makes me curious as to the results of the more common resolutions.

A strong showing at 4K resolution can help with some forms of AA though, including downsampling.

 

[Cloudfire777]

Its actually +31% if you include Sleeping Dogs lol