[Softpedia] Fake GTX-590 Benchmark Results!

[Baasha]

The first results for the NVidia GTX-590 are out!

Nvidia's GTX 590 was tested using a six-core Intel Core i7 980X processor, that was clocked at 3575MHz, together with an un-named Gigabyte X58 motherboard while the Crysis benchmark was run using the Ambush map at a 1920x1080 resolution with 16x anti-aliasing using the Extreme quality profile and the DirectX 10 render path.

The results speak for themselves as the GTX 590 managed to render, in average, 83.61 frames each second while the minimum frame rate recorded was that of 57.59fps.

Screenshots:

Source

Cliffs:

VRAM: 3GB
Cores/Clocks: down-clocked 580s


Editing title per post #70

Moderator Idontcare

 

[busydude]

According to this sub-forum rules.. you need to cite the sources(softpedia in this case) in the title.

OT, looks awesome.

 

[OCGuy]

I'm lazy, can someone link to comparison benchmarks

 

[Baasha]

According to this sub-forum rules.. you need to cite the sources(softpedia in this case) in the title.

OT, looks awesome.

done..

 

[busydude]

Enthusiast shaders with 4XAA

 

[Skurge]

I'm lazy, can someone link to comparison benchmarks

http://www.hardwareheaven.com/revie...s-card-review-crossfire-eyefinity-vs-sli.html

http://www.guru3d.com/article/asus-gtx580-directcu-ii-sli-review/15

 

[RavenSEAL]

So...this thing is gonna cost like...what...$1100?

No thanks...

 

[toyota]

So...this thing is gonna cost like...what...$1100?

No thanks...

its not going to cost any where near that if its a regular production card.

 

[SolMiester]

Interesting, downclocked 580 cores and better AVG FPS?, how does that work?, what am I missing?

 

[OCGuy]

Enthusiast shaders with 4XAA

Is this the exact same map?

OP has 16x AA.

 

[Skurge]

With clocks under 700mhz, this is going to be extremely close with the 6990.

 

[Skurge]

Is this the exact same map?

OP has 16x AA.

No, the AT one should be more demanding, but it has lower settings.

 

[SolMiester]

No, the AT one should be more demanding, but it has lower settings.

Ah...!

 

[hdfxst]

11/17/2010? He ran this benchmark 4 months ago? And gpu-z 0.4.8 was released in november how can it identify a 590?

 

[RavenSEAL]

its not going to cost any where near that if its a regular production card.

Well, based on the fact that GTX580s were going for around $550 at release, i simply just doubled that considering the "590" is supposed to be 2 580 GPUs

 

[notty22]

512 cores x 2 /384 bus .
I guess they could

 

[Skurge]

Ah...!

This is the same level.

http://www.guru3d.com/article/asus-gtx580-directcu-ii-sli-review/15

Look at the last graph.

 

[notty22]

Well, based on the fact that GTX580s were going for around $550 at release, i simply just doubled that considering the "590" is supposed to be 2 580 GPUs

That you would round up 50 bucks x2 and announce you can't afford it ? I can't either

 

[DrBoss]

pretty impressive.

the 590 is rendering 1920x1080 at 16x AA @ 83 avg fps
the 580's SLI'd are rendering 1920x1200 at 2x AA @ 88 avg fps

add some clock to the 590 and you've got a single pci-e slot cranking out the performance of an SLI setup.... hmmmmm, now i want to see some offical 6990 numbers

 

[thilanliyan]

^Lose only 5fps going from 2AA to 16AA with 2 lower clocked 580s on one PCB? I find that really hard to believe.

 

[SolMiester]

This is the same level.

http://www.guru3d.com/article/asus-gtx580-directcu-ii-sli-review/15

Look at the last graph.

Are those AVG FPS?, so same results?, 580SLI = 590.....have to say the scaling doesnt look that flash!

 

[DrBoss]

^Lose only 5fps going from 2AA to 16AA with 2 lower clocked 580s on one PCB? I find that really hard to believe.

the SLI'd 580's are also rendering a higher resolution. Not by much, but 1920x1080 is a decent amount of pixels less than 1920x1200.

 

[OCGuy]

Are those AVG FPS?, so same results?, 580SLI = 590.....have to say the scaling doesnt look that flash!

Uhh....why would you expect scaling to be different just because the chips are on one PCB?

It would make sense that it is the same.

 

[bryanW1995]

what cpu was used for the sli'd 580's? crysis, in my experience, is at least partially cpu-bound even at very high detail levels. let's see some dx11 benchmarks!

 

[SolMiester]

Uhh....why would you expect scaling to be different just because the chips are on one PCB?

It would make sense that it is the same.

I didnt, just poorly worded comment....the scaling looked poor period....however, guess that could be CPU bottleneck there!