ASUS GeForce GTX 590 3 GB

[Frick]

So a driver failure means the card itself is badly designed?

 

[qubit]

So a driver failure means the card itself is badly designed?

Ya know, the more I see people make their points about this card and what it should tolerate, the less I'm sure whether or not it has a hardware design flaw.

Both newtekie1 and mailman78 are maiking good arguments, so it's hard to say.

It seems to me that it's a combination of hardware and software factors that come together to make this go pop. Also, it seems that Asus's claim on the box is in error. nvidia themselves say not to push this much voltage at it (see my earlier post) and it rock solid at it's stock voltage. Also, it could possibly have more phases on the regulator, given how much those power those GPUs pull, further muddying the water.

And I tell ya what, my shiny new GTX 580 has the same power circuitry as the 590 (just x1, obviously) so I ain't never gonna increase the volts on my baby!

 

[CDdude55]

I am curious as to really what caused the problem, you can't just say ''the card is badly designed'', because something had to trigger the card not to protect itself from heavy voltage like it's supposed to, i have heard that some of the drivers being used cuts off the protection so overvolting causes it to screw up, so if that's the case, it's just a matter of swapping drivers.

Or is the power regulation circuitry just not strong enough to handle that amount of voltage?.. or maybe it's cooling, maybe once you slap an aftermarket cooler or waterblock on there then everything is smooth sailing when it comes to heavy overclocks/overvolting(which Nvidia has said recently themselves).

 

[cowie]

Ya know, the more I see people make their points about this card and what it should tolerate, the less I'm sure whether or not it has a hardware design flaw.

Both newtekie1 and mailman78 are maiking good arguments, so it's hard to say.

It seems to me that it's a combination of hardware and software factors that come together to make this go pop. Also, it seems that Asus's claim on the box is in error. nvidia themselves say not to push this much voltage at it (see my earlier post) and it rock solid at it's stock voltage. Also, it could possibly have more phases on the regulator, given how much those power those GPUs pull, further muddying the water.

And I tell ya what, my shiny new GTX 580 has the same power circuitry as the 590 (just x1, obviously) so I ain't never gonna increase the volts on my baby!

no you have 1 more power phase, 590 has 5 per core the 570 4 ..590 sits in between 570 and 580 (6).

i have found that the 6990 had similar issues with some reviewers.... looking for dead 590 cards of all things,but they were to scared to mention it in reviews.
they have good pmw so lets face it the deul cards of both side are trash

http://forums.overclockersclub.com/index.php?showtopic=183386&st=36
BOSCO's posts
page 4-6
Ya the 6990's are not fairing much better I know of 4 cards that are dead so far with 2 more having issues. One of ours died as well..... shakes head

http://www.neoseeker.com/Articles/Hardware/Reviews/AMD_HD_6990_Antilles/18.html


so how did the overvolting go with the 6990 wizz any issues?
i dont see any voltage tweeks in that asus 6990 review maybe not avaliabe at the time? any plans for follow up review with of 20%+ voltage??

 

[qubit]

no you have 1 more power phase, 590 has 5 per core the 570 4 ..590 sits in between 570 and 580 (6).

So, that begs the question of why they chose to use less phases for a card that inherently consumes more, doesn't it? Are the phases more potent than on the 580 perhaps, so they can use less of them? Did they cheap out? Was space too tight on the circuit board? (unlikely). I think it's a good question.

At least I can rest a bit more easily about my card, hehe.

 

[damric]

http://img.techpowerup.org/110326/btajsjbc874d.jpg

is that the easter egg ???

That shit is hilarious.

 

[DaedalusHelios]

Thats irrelevant. They still have to provide proof making such a claim or face false advertisement charges. Even if the circumstanse are wacky like they tested it on the surface of Pluto. They still have to provide proof if questioned.

You think Kingpin or the like won't be doing just that with nicely binned chips? They always do but I know I couldn't get their results even if I tried. Kind of like the theoretical bandwith of things like SATA and USB 2.0 that nobody actually achieves yet are advertised like that is what you can get. The PC market is used to wild claims over the years without proof. Remember all those DX10 and DX10.1 compatible cards that are super low end and cannot handle the DX10 games yet are marketed for "extreme DX10 gaming". Do they meet the minimum system requirements even? If marketing was honest we wouldn't need reviewers for the most part.

 

[newtekie1]

So I guess you need to submerge the entire card in LN?

Accroding to even Nvidia "This should not have happen"


Quote:
According to NVIDIA this should not happen. In their official reviewer driver (which I used), the NVIDIA Power limit is designed to be active for all applications, not only Furmark.

So yeah something failed here. It failed in other reviews. Again this is not an isolated incident and the card sucks in its current revision, blown cards across a few sites proves this. Now maybe when the new one W1zz talked about hits the streets you might have something worth buying.

That is a fuse that blew when the mosfet failed, your point?

And again, the power limitter won't work at 1.2v, and nVidia says you should never go that high. What part of that don't you understand?

I haven't seen a single other reivew that blew the card at or below the nVidia set maximum of 1.06v, have you? The exception being Sweclockers, which blew their card because they were using drivers that didn't have the power limitter in place.

What is interesting is that I just read another review that used smartdoctor, and went a little bit more in depth about smartdoctor with this card. It seems that smartdoctor actually does bypass the BIOS limit on the card, allowing voltages up to 1.213v. In this case, it seems ASUS' software is to blame for bypassing that limit without even a warning to the user. Also, the same review actually contacted nVidia before just jacking up the voltages and clocks and here is what nVidia told them:

It is not advised to generally exceed the 1.050 to 1.065 vcore range as this begins to meet the limits for the OCP/OVP mechanism on the card. Exceeding these values without disabling OCP/OVP or having superior cooling could affect the lifespan and functionality of the card/gpu.

I don't really feal like this is going anywhere anymore, so I'm done discussing it.

 

[sLowEnd]

http://img.techpowerup.org/110327/package1.jpg

FYI that is a generic logo they use for all their recent cards that support overvolting.

...etc

If you read the fine print, it also mentions that it's "up to 50%"

 

[thedude74]

I am curious as to really what caused the problem, you can't just say ''the card is badly designed'', because something had to trigger the card not to protect itself from heavy voltage like it's supposed to, i have heard that some of the drivers being used cuts off the protection so overvolting causes it to screw up, so if that's the case, it's just a matter of swapping drivers.

Or is the power regulation circuitry just not strong enough to handle that amount of voltage?.. or maybe it's cooling, maybe once you slap an aftermarket cooler or waterblock on there then everything is smooth sailing when it comes to heavy overclocks/overvolting(which Nvidia has said recently themselves).

Nvidia said this:

The few press reports on GTX 590 boards dying were caused by overvoltaging to unsafe levels (as high as 1.2V vs. default voltage of 0.91 to 0.96V), and using older drivers that have lower levels of overcurrent protection.

Which basically says it all.

Additionally, anyone still clinging to the idea that the protection should have kicked in at these ridiculous voltages should also read this.

Please note that overcurrent protection does not eliminate the risks of overclocking, and hardware damage is possible, particularly when overvoltaging.

Which basically means, you run the card in unintended ways you can end up with unintended consequences.


Now, when a rash of 590's burn up at stock voltages or even overclocked voltages that are within design parameters, I'll start agreeing with people calling this card junk, until then they don't have a leg to stand on.

The 590 in the review was clocked at 612mhz stock with a voltage of .938v. At 1.0v it hit 815mhz with stock cooling! That's a 33% overclock and is faster than a 580! What a horrible card!

Finally, where is the similar outrage in terms of the 6900, which as far as I can tell is the ONLY one of these two cards that is failing under proper use? Why no mention of these cards dying and why is it a big secret? Why no crazy overvolt test's to see how far a 6990 could be pushed? Seems like a double standard to me.

 

[gaximodo]

The cards is quieter, smaller, and as strong as the HD6990 for most of us (1920*1200/1080) while providing more than twice of overclocking potential at stock voltages, at the same price.

wrong thread.. OOOPS

 

[cowie]

Nvidia said this:



Which basically says it all.

Additionally, anyone still clinging to the idea that the protection should have kicked in at these ridiculous voltages should also read this.



Which basically means, you run the card in unintended ways you can end up with unintended consequences.


Now, when a rash of 590's burn up at stock voltages or even overclocked voltages that are within design parameters, I'll start agreeing with people calling this card junk, until then they don't have a leg to stand on.

The 590 in the review was clocked at 612mhz stock with a voltage of .938v. At 1.0v it hit 815mhz with stock cooling! That's a 33% overclock and is faster than a 580! What a horrible card!

Finally, where is the similar outrage in terms of the 6900, which as far as I can tell is the ONLY one of these two cards that is failing under proper use? Why no mention of these cards dying and why is it a big secret? Why no crazy overvolt test's to see how far a 6990 could be pushed? Seems like a double standard to me.

They cant be honest about any amd card short comings...amd will not send them any more review cards.
we should let karma run its coarse it will came back to bite in the ass somehow someway
Maybe in a month or so time when/if these 6990 start pile'in up in droves the unwarned public will wake up on the amd card reviews and be more of a skeptic or just stop reading them altogether.
remember only two brands will not void warrenty if the bios switch is moved!!! that alone is enuogh for me to only recomend two brands of that card.

 

[Harlequin_uk]

http://www.nvidia.com/object/win7-winvista-64bit-267.91-driver.html

new driver out - guessing its fixed now

 

[the54thvoid]

When a Tweak isn't a Tweak.

The actual word 'tweak' means to 'fine tune'. A 25% adjustment in voltage isn't a tweak. So if you want to go on semantics, what some reviewers have done is voltage 'boost'.

The Hardware Canucks article also has this paragraph:

In addition, here is an extract from ASUS’ GTX 590 reviewer’s guide:

It is not advised to exceed the 1.050 to 1.065 vcore range as this begins to meet the limits for the OCP/OVP mechanism on the card. Exceeding these values without disabling OCP/OVP or having superior cooling could affect the lifespan and functionality of the card/gpu.

I have read through the posts so i don't think this has been posted up but it should point as an indicator that if cards fried, it's because even our beloved W1zz wanted to push it (unless that's what the e-mails were all about but my reading was that was an NV email not the actual ASUS reviewer guide.?).

Other review sites have pushed the clock up and didn't bother overvolting. Reviewing the reviews, most that OC'd got 690 core, even without voltage changes.

The argument shouldn't be about fried cards. The cards fried due to
a) wrong drivers,
b) in the ASUS case, reviewers not following guidelines
c) MOST IMPORTANTLY, Nvidia failing to tell the reviewers exactly what drivers to use or not having them on time.

The card doesn't suck. In almost all reviews it gets praise for what it delivers, performance without the associated drawbacks of noise and size. The fact some folk killed it is NV's fault for poor marketing and technical advice.

Dare I say it W1zzard's review is one of the most scathing. Most places give it 8 or 9 out of 10.

If the new driver doesn't allow the same cock ups (and no more cards fry) then this whole argument is irrelevant.

 

[Mussels]

reviewers not following guidelines? those guidelines were given out AFTER the reviewers ended up with dead cards, and informed everyone of the problem.

the rest of your post i agree with.

 

[TheMailMan78]

reviewers not following guidelines? those guidelines were given out AFTER the reviewers ended up with dead cards, and informed everyone of the problem.

the rest of your post i agree with.

Its like the reviewers were the beta testers. Sad day for Nvidia.

 

[Harlequin_uk]

well so far theres an increasing amount of dead 6990`s as well.


too much from both `teams`.

 

[cadaveca]

well so far theres an increasing amount of dead 6990`s as well.


too much from both `teams`.

Um, links?

 

[Harlequin_uk]

Um, links?

scroll up?


http://www.techpowerup.com/forums/showpost.php?p=2238253&postcount=479

 

[CDdude55]

well so far theres an increasing amount of dead 6990`s as well.


too much from both `teams`.

I have not seen any dead 6990's from any reviews, so that's a lie unless i see some links. The only issues ive seen with the 6990 is that it's loud as fuck and super expensive. (and performance is very similar to the GTX 590)

Its like the reviewers were the beta testers. Sad day for Nvidia.

The reviewers are always the ''beta testers'', and of course as thoroughly explained before, i find it hard to blame Nvidia for such a non common issue. I guarantee 98% of people buying this card won't have the same problem, as there are different standards for reviews and consumers. Those that what to overvolt know the risks on such a card, but it's far from a mainstream problem.

 

[cadaveca]

scroll up?


http://www.techpowerup.com/forums/showpost.php?p=2238253&postcount=479

I have not seen any dead 6990's from any reviews.

Neoseeker, apparantly.

After our initial testing of the HD 6990 we moved the graphics card over to a backup system that we were using to test new games for our benchmarking suite. We were able to complete testing with the HD 6990 in some of our new benchmarks including H.A.W.X 2, Lost Planet 2 and DiRT 2, however, when we were testing the performance of Dragon Age II the HD 6990 died on us. At the time of it's demise the card was set at the stock 830MHz setting and the BIOS switch was in the default position. The fact that it died could have been that we tested the graphics card at both the 375W and 450W settings, but since the review we have left the settings at default level.

 

[wahdangun]

The actual word 'tweak' means to 'fine tune'. A 25% adjustment in voltage isn't a tweak. So if you want to go on semantics, what some reviewers have done is voltage 'boost'.

The Hardware Canucks article also has this paragraph:



I have read through the posts so i don't think this has been posted up but it should point as an indicator that if cards fried, it's because even our beloved W1zz wanted to push it (unless that's what the e-mails were all about but my reading was that was an NV email not the actual ASUS reviewer guide.?).

Other review sites have pushed the clock up and didn't bother overvolting. Reviewing the reviews, most that OC'd got 690 core, even without voltage changes.

The argument shouldn't be about fried cards. The cards fried due to
a) wrong drivers,
b) in the ASUS case, reviewers not following guidelines
c) MOST IMPORTANTLY, Nvidia failing to tell the reviewers exactly what drivers to use or not having them on time.

The card doesn't suck. In almost all reviews it gets praise for what it delivers, performance without the associated drawbacks of noise and size. The fact some folk killed it is NV's fault for poor marketing and technical advice.

Dare I say it W1zzard's review is one of the most scathing. Most places give it 8 or 9 out of 10.

If the new driver doesn't allow the same cock ups (and no more cards fry) then this whole argument is irrelevant.

yes it is some card dead with default clock and voltage

^^ This , i know already of 5 dead cards from my friends that died at stock voltages stock clock running heaven and Vantage benchmark , add 2 more testing on LN2 and 2 more on H2O = 9 cards already , how many dead 590s have not been reported ? this card cannot even handle 100% full load on GPU for long periods of time , yes it will blow up in smoke even at stock voltages and stock clocks. Nvidia should make a recall on this cards me thinks :/ , so many cards 590s have died and since has not been reported it looks thats just a few cards , the truth is , its going to happen soon or later , its like a time bomb , this card its a total failure and im sad i had high expectations on this card. Ill wait for the next 28mm cards to be release , as of right now 590 or 6990 are not apealing to me at all

linky

chipsy is a member of HWbot, if you don't trust me just go to guru3d forum and ask everyone about chipsy.

 

[EarthDog]

Question for your Wizzard... And I didnt check this thread for this answer...

Did you intentionally put EPIC FAIL in your 'thumbs down' section of your review summary? (first letter of each sentence)

 

[the54thvoid]

reviewers not following guidelines? those guidelines were given out AFTER the reviewers ended up with dead cards, and informed everyone of the problem.

the rest of your post i agree with.

Its like the reviewers were the beta testers. Sad day for Nvidia.

Yeah, that's kinda my point - NV do know how to massively cock things up. How much better would the PR be if they waited a few more days for the appropriate guidance/testing or drivers?

 

[TheMailMan78]

yes it is some card dead with default clock and voltage




linky

chipsy is a member of HWbot, if you don't trust me just go to guru3d forum and ask everyone about chipsy.

Most of these Nvidia users don't even want to believe W1zz. Do you think they will believe this guy?

Moderators---> <--- Mailman.