[VR-Zone] NVidia GTX-590 *FINAL* Specs Revealed!

[Wag]

On sale already? http://www.expc.ca/i-17979-asus_engtx590_3dis_3gd5.html

Check out the box.

 

[3DVagabond]

$799! Does that mean it's faster than the 6990? Or, is this pre-release gouging? Or, does JHH have the brass to charge more for a slower card because it might O/C to be faster?

I'm thinking it might actually be faster. Although @ 612 clocks, I don't see how it could be.

 

[MentalIlness]

Item Description

Model
ENGTX590/3DIS/3GD5
Graphics engine
NVIDIA® GeForce® GTX 590
Bus standard
PCI Express® 2.0

Video memory
3GB GDDR5

Engine clock
612MHz

CUDA cores
512*2

Memory clock
3420MHz (855MHz GDDR5)

Memory interface
384-bit x2

DVI max resolution
2560 x 1600
D-Sub max resolution
2048 x 1536
D-Sub output
YES X1 (via DVI-to-D-Sub adapter)
DVI output
Native dual-link DVI-I X3
HDCP compliant
Yes
HDMI output
YES X1 (via DVI-to-HDMI adapter)
DisplayPort output
Mini DisplayPort X1
Accessories bundled
DVI-to-HDMI adapter x 1
DVI-to-D-Sub adapter x 1
Power cable
Software bundled
ASUS utilities and driver
*Follow driver setup instructions to download the SmartDoctor application from www.asus.com prior to use
Dimensions
11” x 4.5”

 

[dakU7]

Voltage tweak! 50% faster?

 

[MentalIlness]

Voltage tweak! 50% faster?

I highly doubt it will run 50% faster. I just don't see a voltage tweak "if your running 100 fps" increasing the performance to "150 fps".

Unless I completely figured that wrong.

 

[facepalm]

can anyone confirm dat when HD6990 prices showed up on early stages what was the preorder price?

 

[TerabyteX]

If those are its final specs, I don't see how it will outperform the HD 6990, plus if its says 50% more performance through voltage tweaks, I really doubt somehow its overclockability potential as the GTX 580 wasn't a spectacular overclocker anyways and even if you aim and bin the best GTX 580 dies from the waffer, its rare to see reductions greater than 15% in terms of power tweaks like lower Vcore and power consumption.

 

[3DVagabond]

Voltage tweak! 50% faster?

Never noticed that before? That's Asus' standard propaganda.

 

[dakU7]

So looks like 590 is going to have a bios switch.. wouldn't that overload the PCI in terms of voltage?

 

[MrK6]

Looks like AMD's power efficiency might give them the performance crown this round. It'll be interesting to see what full reviews come up with. I can't say I'm impressed with either card though.

 

[badb0y]

It's obvious the die shrink was needed.

 

[Baasha]

Want to see EVGA's version of the GTX-590.

Also, there is a rumor, actually, a tweet from an EVGA employee, that there is a 3GB version of the GTX-580 they're making! Wonder if EVGA will make a GTX-590 version based on that (6GB total). Now, THAT, would be insane!

So the waiting game continues.

If it is between 2 GTX-590s and 3 GTX-580 (3GB), which one would be the better choice for ultra-high resolutions and surround gaming?

 

[MangoX]

Wonder how much that 6GB 590 gonna cost... BTW why didn't Nvidia name it as 595? Are they gonna release a 595 with stock clocked 580's and 3x 8pin power?

 

[Castiel]

Check out the box.

Edit.. Nevermind

 

[Madcatatlas]

The GPU itself would be outdated or left in the dust when that amount of memory was "needed". And professional cards already have much more ram.

 

[Silverforce11]

Considering SLI 580s at those speeds already lose to the 6990 at 1080p res (and still using a good 100w more), it's not going to be pretty, especially moving towards 30' res or triple monitors where these cards are aimed at.

Next battle at 28nm then.

 

[Idontcare]

Looks like AMD's power efficiency might give them the performance crown this round. It'll be interesting to see what full reviews come up with. I can't say I'm impressed with either card though.

If you think about it, that has more or less been the name of the game since 90nm, be it for CPU's or GPU's.

I'm dissapointed that AMD isn't migrating their GPU's to the same high-performance HK/MG SOI process tech that their CPU's benefit from.

An architecture advantage combined with a process tech advantage could have resulted in some rather intriguing products being brought to market.

 

[bryanW1995]

The GPU itself would be outdated or left in the dust when that amount of memory was "needed". And professional cards already have much more ram.

then why does 6970 approach gtx 580 as you get to higher and higher resolution with higher and higher detail settings? 3gb might be overkill for a single card, but for 3-4 of them at very high resolution/multi monitor settings it would probably make a significant difference.

 

[Kenmitch]

If you think about it, that has more or less been the name of the game since 90nm, be it for CPU's or GPU's.

I'm dissapointed that AMD isn't migrating their GPU's to the same high-performance HK/MG SOI process tech that their CPU's benefit from.

An architecture advantage combined with a process tech advantage could have resulted in some rather intriguing products being brought to market.

This release from AMD and Nvidia is more of a practice run....Keep the wheels turning kinda thing!

I think it'll be interesting to see what both of them can do in the next round or two once they aren't confined to the limits of 40nm!

 

[MrK6]

If you think about it, that has more or less been the name of the game since 90nm, be it for CPU's or GPU's.

I'm dissapointed that AMD isn't migrating their GPU's to the same high-performance HK/MG SOI process tech that their CPU's benefit from.

An architecture advantage combined with a process tech advantage could have resulted in some rather intriguing products being brought to market.

Very interesting. Forgive my ignorance on the entire subject, but is this process tech something Global Foundries will employ, and could we look forward to it with 28nm? Also, what specific tweaks/benefits does the HK/MG SOI process afford? Lower power consumption? Higher clocks?

then why does 6970 approach gtx 580 as you get to higher and higher resolution with higher and higher detail settings? 3gb might be overkill for a single card, but for 3-4 of them at very high resolution/multi monitor settings it would probably make a significant difference.

AMD's cards have been better at higher resolutions for awhile now. It's not usually a RAM capacity issue - if it was, you'd see the the GTX 580 fall hard on it's face as the card has to page. However, it generally doesn't, as it simply just doesn't perform as well. I remember it being mentioned that AMD has a better memory management sub-system, but I'm sure there's more to it.

 

[tviceman]

AMD's cards have been better at higher resolutions for awhile now. It's not usually a RAM capacity issue - if it was, you'd see the the GTX 580 fall hard on it's face as the card has to page. However, it generally doesn't, as it simply just doesn't perform as well. I remember it being mentioned that AMD has a better memory management sub-system, but I'm sure there's more to it.

Isn't it (in situations where 1.5 gigs of ram is more than enough) just a straight up fillrate issue? AMD cards have a higher fillrate than Nvidia cards, right?

 

[SlowSpyder]

then why does 6970 approach gtx 580 as you get to higher and higher resolution with higher and higher detail settings? 3gb might be overkill for a single card, but for 3-4 of them at very high resolution/multi monitor settings it would probably make a significant difference.

Does it have anything to do with TMU power? I don't know how that might play into it, but I know AMD cards have significantly more ability in that area.

 

[Arkadrel]

GPU clocks: 772 -> 612 (27% downclock)
MEM clocks: 1002 -> 855 (18% downclock)

I seriously doubt this will be faster than the 6990... so it ll cost more (590's 800$+ vs 700$ of the 6990) and be slower, wonder how review sites will "sell it", reguardless enough nvidia only buyers out there, that it ll still sell.

 

[AtenRa]

Does it have anything to do with TMU power? I don't know how that might play into it, but I know AMD cards have significantly more ability in that area.

I don't think is Texture related but it could have to do more with ROPs (Render Back End).
Cayman ROPs have been upgraded although they have been kept at the same number of 32.

 

[bryanW1995]

hmmm, it still seems reasonable that 3gb per card would come in handy on a multi monitor tri-sli setup though, right?