Official GTX 590 Review Thread (23 reviews at this time)

[pcm81]

link..? I seen the video and it did not state voltage.
Would you now admit to sticking 1.2 volts into a card? I know I wouldn't, I wouldn't want to be at the retard reviewers table.

And even if it did, thats one of thousands, and it happends.

Cant we all just get along.

I think it will make every one happy if we all just admit that 590 is a bigger BANG for the buck... (Pun intended)

 

[Firestorm007]

Fixed that for you.

nm

 

[Outrage]

link..? I seen the video and it did not state voltage.
Would you now admit to sticking 1.2 volts into a card? I know I wouldn't, I wouldn't want to be at the retard reviewers table.

And even if it did, thats one of thousands, and it happends.

it's in the description of the video http://www.youtube.com/watch?v=sRo-1VFMcbc

And i wouldn't call W1zzard a retard, he and other reviewers have probably punished other cards in the same way without them going up in smoke. the gpu/memory starts showing signs that they are at there limit long before the card explode in your face. This card is clearly to weak of a design to handle 2 gf110 gpu's.

edit: and they killed 2 cards at those settings (1,025v), they tried it again at the request of a card supplier. so the 1 in 1000 is bs.

 

[insurgent]

link..? I seen the video and it did not state voltage.
Would you now admit to sticking 1.2 volts into a card? I know I wouldn't, I wouldn't want to be at the retard reviewers table.

And even if it did, thats one of thousands, and it happends.

Wow, calling someone who has done so much for the enthusiast community a retard... He overclocks and overvolts the cards he reviews, that's what he does. So what really caused the card dying? It's the stupid nvidia driver that came with the card, he overvolted it knowing that OCP will protect the card but it didn't, so much for that "engineering marvel", more like a POS.

 

[hdfxst]

I think in all of W1zzard reviews he pushes the voltage as far as the sliders will go.In the powercolor 6970 review he put over 1.4 volts to it

 

[MrK6]

I love the fact that reviewer's are called "retarded" for finding a fault in hardware. Rampant fanboyism really is taking down this forum.

 

[Elfear]

I think in all of W1zzard reviews he pushes the voltage as far as the sliders will go.In the powercolor 6970 review he put over 1.4 volts to it
Plus maybe now people will think twice before blindly moving sliders and hitting apply

I'm sure W1zzard was just blindly moving sliders. The guy does a lot for the enthusiast community and has been overclocking for a long time. Somehow I don't get the picture that he's like Homer in the nuclear plant.

 

[hdfxst]

I'm sure W1zzard was just blindly moving sliders. The guy does a lot for the enthusiast community and has been overclocking for a long time. Somehow I don't get the picture that he's like Homer in the nuclear plant.

That's not how i meant it i wasn't trying to bash W1zzard at all.I meant it for people who buy a card expecting it to overclock and what can go wrong.
I edited it out to avoid confusion

 

[TerabyteX]

For a marvel of engineering flagship product, this thing sure is fragile.

May be their duct tape wasn't good enough to hold it together.

 

[ViRGE]

The different scaling they used was just to show more easily the differences between several parts of the card. The maximum temperature they are reporting is correct, regardless of which scale they use. Excuse this post if that is what you meant in what I quoted.

Well yeah, that was half my point. My primary point is that the whole card isn't as hot as those pictures indicate because the range used can't show 90C-120C properly. I'm sure there are hotspots under the VRMs, I just doubt the rest of the card is that hot.

However my secondary point is that I doubt that spot is 111.9, particularly since it's outside the camera's range and that he got that value twice. It's over 100C I'm sure, I just doubt it's precisely 111.9.

 

[Firestorm007]

I love the fact that reviewer's are called "retarded" for finding a fault in hardware. Rampant fanboyism really is taking down this forum.

Agreed. The fact of the matter is that the 6990 is faster...bar none...It may offend some, but if you're truly a gamer, red or green doesn't really matter. However, the "retarted" comment is way out in left field. If you are so in love with a particular vendor for whatever reason, so be it. But, don't lose sight of reality and the evidence right in front of you. It just makes you look petty.

 

[thilanliyan]

However my secondary point is that I doubt that spot is 111.9, particularly since it's outside the camera's range and that he got that value twice. It's over 100C I'm sure, I just doubt it's precisely 111.9.

I don't think he got that value twice. He just changed the scale for the 2nd image.

Did he say he took two measurements? My French isn't great.

 

[pcm81]

according to gpu-z my 6990 peaks at 110C and his is at stock 830MHz clock 100% load running boinc... I wonder if boinc actually stresses it more than 3d mark... btw ambient in the room is 80F so that dont help either...

EDIT: I have 2 cards side by side, so fan intake is suffering on #1 card...
Idle its at 60. 2nd card Idle is at 50C.

 

[bryanW1995]

Sweclockers card exploded @ 1,025v

quit confusing him with "facts".

 

[bryanW1995]

according to gpu-z my 6990 peaks at 110C and his is at stock 830MHz clock 100% load running boinc... I wonder if boinc actually stresses it more than 3d mark... btw ambient in the room is 80F so that dont help either...

EDIT: I have 2 cards side by side, so fan intake is suffering on #1 card...

different boinc programs can stress the card more. Milkyway@home is a good program for ppd on amd cards, but I don't know if that pushes it to 100% or not.

 

[pcm81]

different boinc programs can stress the card more. Milkyway@home is a good program for ppd on amd cards, but I don't know if that pushes it to 100% or not.

I use milkyway and collatz. GPUgrid now also works on 6990, did not on 4870x2.

 

[Baasha]

Somebody suggested extending the PCB of the GTX-590 and adding more VRMs might have solved this issue. I thought that was a perspicacious comment as NVidia was trying to claim 'quiet', 'powerful', AND '(somewhat) small' for the GTX-590. Too ambitious IMO. Someone who spends $700+ on a GPU usually has ample space in their cases. If the GTX-590 was as long as the AMD 6990, the voltage issue would not be such a concern.

What do you guys think?

 

[Kenmitch]

Somebody suggested extending the PCB of the GTX-590 and adding more VRMs might have solved this issue. I thought that was a perspicacious comment as NVidia was trying to claim 'quiet', 'powerful', AND '(somewhat) small' for the GTX-590. Too ambitious IMO. Someone who spends $700+ on a GPU usually has ample space in their cases. If the GTX-590 was as long as the AMD 6990, the voltage issue would not be such a concern.

What do you guys think?

Sounds logical to me

 

[3DVagabond]

it's in the description of the video http://www.youtube.com/watch?v=sRo-1VFMcbc

And i wouldn't call W1zzard a retard, he and other reviewers have probably punished other cards in the same way without them going up in smoke. the gpu/memory starts showing signs that they are at there limit long before the card explode in your face. This card is clearly to weak of a design to handle 2 gf110 gpu's.

edit: and they killed 2 cards at those settings (1,025v), they tried it again at the request of a card supplier. so the 1 in 1000 is bs.

Impossible!!!! All 590's can handle 1.05-1.1 and O/C to 800MHz!!! Pay no attention to the facts that reviewers have put in black and white. [/sarc]

Seriously, this is not the first of nVidia's new generation that have had this problem. It's just the latest.

@happymedium, I notice that you didn't buy a reference 460, or a Palit, or some other cut rate bargain design. You bought the MSI Hawk. The 460 with the most over engineered power stage and one of the best coolers on the market. Why don't you just be honest with people and tell them to hold off until there's a Hawk or equivalent model available. We all know the potential of the 590. The reference model just doesn't fulfill that potential.

This is assuming the OCP in the drivers doesn't hold it back. Which I'm concerned it might (dumb way of doing it).

 

[pcm81]

Sounds logical to me

There is no magic in real life, you gotta loose something to make something smaller...

 

[1h4x4s3x]

I love the fact that reviewer's are called "retarded" for finding a fault in hardware. Rampant fanboyism really is taking down this forum.

Yup. That's because there are no consequences to your actions (trolling, spreading fud etc.).
Funnily enough, I got an infraction for asking what notty's motivation is on page 1.

Anyway, it's 'only' a driver issue, isn't it?

 

[3DVagabond]

http://www.guru3d.com/article/gigabyte-geforce-gtx-590-review/15

** We do want to make a note here -- a previous BETA driver for the GTX 590 showed much better results then the beta for this review, which was 267.71 So I expect in a final driver the performance to jump back to normal again. But we report what we measure, it is as simple as that.

Just found this in Guru3D's latest 590 review (Gigabyte). Looks like, in crysis at least, the original release drivers were higher performance than the later ones. Someone else speculated that maybe nVidia pushed it too far with those original drivers trying to compete with the 6990. This doesn't prove it, but it does supply some evidence to that theory. I wonder how many of the benches were run with the original drivers and not redone?

 

[happy medium]

@happymedium, I notice that you didn't buy a reference 460, or a Palit, or some other cut rate bargain design. You bought the MSI Hawk. The 460 with the most over engineered power stage and one of the best coolers on the market.

I bought the card from a freind. It was a good chip that I knew I could get to the 1ghz mark with.
I don't care if god engineered my cards's power stage and it had a 10,000 watt air conditioner attached to it, I STILL WOULD NOT PUMP 1.2 VOLTS INTO IT, ESPECIALLY IF IT WAS A BRAND NEW, NEWLY DESIGNED, 700$ DUAL gpu CARD.
ONLY RETARDS DO THAT WITHOUT WATERCOOLING.

 

[badb0y]

I bought the card from a freind. It was a good chip that I knew I could get to the 1ghz mark with.
I don't care if god engineered my cards's power stage and it had a 10,000 watt air conditioner attached to it, I STILL WOULD NOT PUMP 1.2 VOLTS INTO IT, ESPECIALLY IF IT WAS A BRAND NEW, NEWLY DESIGNED, 700$ DUAL gpu CARD.
ONLY RETARDS DO THAT WITHOUT WATERCOOLING.

They are dying from various voltages not just 1.2. I would say anything over 1.00v is unsafe for the card at this time.

 

[3DVagabond]

I bought the card from a freind. It was a good chip that I knew I could get to the 1ghz mark with.
I don't care if god engineered my cards's power stage and it had a 10,000 watt air conditioner attached to it, I STILL WOULD NOT PUMP 1.2 VOLTS INTO IT, ESPECIALLY IF IT WAS A BRAND NEW, NEWLY DESIGNED, 700$ DUAL gpu CARD.
ONLY RETARDS DO THAT WITHOUT WATERCOOLING.

WTF? Well, I guess your lucky that your friend knows how to pick a well engineered model then.