SweClockers: Geforce GTX 590 burns @ 772MHz & 1.025V

[Timorous]

This is probably up there with the FX5800 in terms of poor launches. The fact that it takes drivers to protect this card is really unnerving because if the company can let the card slip through its QA in this state what makes us think they wont release a driver that has an error in the over current protection system causing them to fry? They have released killer drivers before so it is not without precedent.

I also think if this card is downclocking itself in normal usage conditions in a cool climate like the UK then there could be an argument made that this product is in breach of the Sales of Goods Act and the Trade Descriptions Act (UK sales statutes that ensure you are buying a product that is as described, is fit for purpose, is of sufficient quality and has a reasonable life expectancy).

Downclocking to protect from heat or over current damage is fine but these protections should only kick in if you are overclocking heavily which puts the card out of spec or if there is a failure of the cooling device in some way. It should not be occurring during standard operating conditions if you happen to live in a hot climate because the device should be designed with that in mind.

 

[Silverforce11]

Peak wattage is not 350W, that’s the wattage when we have 1.0V (watt = Amps * Volts). Default voltage of GTX590 is 0.938V meaning that the cores will draw (10 x 35A) * 0.938V = 328,3W.

If we over-voltage to 1.05V then the two cores will draw 367,5W. Remember that this is only for the two GF110 chips and not for the entire card and this is only the theoretical maximum Wattage because the MOSFET has less than 100% efficiency.

The Infineon MOSFET (TDA21211) can supply the necessary wattage to the two CF110 chips so something else is happening.

Edit: I could be wrong but IMHO I will say its heat related.

From the specs and graphs, its around ~85% efficiency with peak at around 1V. Thus i rated them ~350W max. Remove some for conversion loss and you get less. Thats why at default the VRMs are already under stress. NV dropping clock speed at max load with the recent "safe" drivers is to account for this problem.

Btw, at default clocks under load 590 pulls over 400W already from quite a few reviews. It's consistently higher than the 6990 (~50W higher) which we also know will max out at around 370W at default.

I don't think its a heat issue since ppl have measured VRM temps under load and its ~100C which is fine.

 

[taltamir]

You see an organized smear campaign, I see fanboys and enthusiasts doing what they've always done - overhype and over-dramatize a video card launch, yet again, on their Mt Dew/redbull fueled benders.

It is one big drama fest. So far the claims of "dying at stock" ended up being "I know ofs" or probably lies (melted through tungsten!). You supposition that it is a bunch of fans and enthusiasts getting caught in the drama rather then an organized smear campaign is rather plausible; Humans are prone to forming angry mobs at the drop of a hat and for all the wrong reasons.

But as over dramatized as it is, there seems to be a kernel of truth; the GTX590 does have issues.

 

[WelshBloke]

It is one big drama fest. So far the claims of "dying at stock" ended up being "I know ofs" or probably lies (melted through tungsten!). You supposition that it is a bunch of fans and enthusiasts getting caught in the drama rather then an organized smear campaign is rather plausible; Humans are prone to forming angry mobs at the drop of a hat and for all the wrong reasons.

But as over dramatized as it is, there seems to be a kernel of truth; the GTX590 does have issues.

He was joking, it was pretty obvious. I assume you're referring to this quote?

I down clocked my GTX590 to 4MHz on the cores, it fried and melted through my custom tungsten case while I was playing Minesweeper!!!!!1!!!2!!

 

[taltamir]

He was joking, it was pretty obvious. I assume you're referring to this quote?

Eh, you can never tell on the internet. But I see why you would think it was a joke, it does sound like it would be sarcasm.

 

[3DVagabond]

Eh, you can never tell on the internet. But I see why you would think it was a joke, it does sound like it would be sarcasm.

Ya think?

 

[taltamir]

Ya think?

You laugh, but I have seen worse where the person was serious

 

[insurgent]

That tungsten case and thermite hsf bit was funny though.

 

[AtenRa]

From the specs and graphs, its around ~85% efficiency with peak at around 1V. Thus i rated them ~350W max. Remove some for conversion loss and you get less. Thats why at default the VRMs are already under stress. NV dropping clock speed at max load with the recent "safe" drivers is to account for this problem.

Btw, at default clocks under load 590 pulls over 400W already from quite a few reviews. It's consistently higher than the 6990 (~50W higher) which we also know will max out at around 370W at default.

I don't think its a heat issue since ppl have measured VRM temps under load and its ~100C which is fine.

It is strange that Infinion rates the MOSFET at 35A max but the graphs tops at 30A. If the line in the graph continues to drop the efficiency at 35A and 1V will be close to 81-82%.
1.0V is not the maximum those MOSFETs can handle, from the bellow diagram it clearly says that 1.1V is within the operational conditions.

As I have said before those 10 MOSFETs give power to the two GF110 chips only. We have two more (4 in total) VRMs for each memory to contribute power to the cards total Power usage and other logic.
So the 400W is for the entire card and not for the two GF110 chips so the MOSFET can handle that.

Question, can NV lower the Voltage through a new driver ??

Because if the new driver only down-clock the GF110 frequency to 550MHz the Voltage will remain the same and if that’s the case then they only down-clocked it for heat issues.

http://html.alldatasheet.com/html-pdf/312873/INFINEON/TDA21211/5468/14/TDA21211.html

 

[Silverforce11]

What do you mean can they drop Voltage with a downclock? A 57mhz drop will result in less power draw. It won't drop the vcore, it just reduces the power use thus easing off on the VRMs which are under stress. Just as when you OC the core leading to a big jump in power use, lowering clock has the opposite effect. It's not about heat, VRMs can handle ~115C operation conditions.

That is essentially what NV is doing with these recent drivers. If OCP senses the card is maxing out, it just throttles to 500-550mhz to stay safe.

Edit: It's not heat, the fan blows directly onto the VRMs first before cooling the GPUs. Reviews which measured VRM temps, it wasn't extraordinary. For some maths, its ~329W @ 0.94vcore (close enough) * 0.85 efficiency = 279W. Two Fermi cores = 139W per core. That's the problem.

 

[AtenRa]

What do you mean can they drop Voltage with a downclock? A 57mhz drop will result in less power draw. It won't drop the vcore, it just reduces the power use thus easing off on the VRMs which are under stress. Just as when you OC the core leading to a big jump in power use, lowering clock has the opposite effect. It's not about heat, VRMs can handle ~115C operation conditions.

That is essentially what NV is doing with these recent drivers. If OCP senses the card is maxing out, it just throttles to 500-550mhz to stay safe.

Edit: It's not heat, the fan blows directly onto the VRMs first before cooling the GPUs. Reviews which measured VRM temps, it wasn't extraordinary. For some maths, its ~329W @ 0.94vcore (close enough) * 0.85 efficiency = 279W. Two Fermi cores = 139W per core. That's the problem.

If you lower the GF110 chip frequency to 550MHz it will lower the power the GF110 will output (though lowering the Total card power output) PUT the MOSFET will continue to operate at 35A and 0.938V the same conditions as before.

If the problem was on the MOSFETs then they would have to lower the Voltage.

LAB501 managed to OC the card to 900MHz (1.1V) with custom AIR cooling and that’s shows that the VRMs are capable to deliver the Voltage the two cores needs at 900MHz.

http://lab501.ro/placi-video/nvidia-geforce-gtx-590-studiu-de-overclocking/7

Edit: Lowering the power the GF110 chips will output will lower the heat they generate.

 

[Soleron]

I don't know tbh, but it seems rather plausible and logical that it would throttle like that.

Intel cpu's have done that since Nehalem, get'em too hot and drawing too many amps and they'll underclock themselves too.

The article is suggesting it does that under normal gaming loads. No other GPU or CPU I've ever heard of does that.

If true, this was planned. Get the reviewers to review at an unsustainable power draw then quietly downclock in the next driver release.

 

[Silverforce11]

How come you seem to think the VRMs deliver or operate the same regardless of load? If the GPUs clocks are lower, they will draw less power.. the VRMs do not continue to feed it with max current loads. If that were the case, idle clocks wouldn't affect power use and subsequently cooling down the VRMs or reducing the current drawn (you can check VRM amps).

ICs run more efficient at cooler temps, obviously these guys put some massive heatsinks on the VRM and beefy coolers on it allow it to run well beyond recommended specs.

In the OC review you linked: http://lab501.ro/placi-video/nvidia-geforce-gtx-590-studiu-de-overclocking/12
The reviewer specifically said the 590 should have had 6 Volterra 40A VRMs per core rather than the 5x bad TDA 35A with low efficiency. It limited him pushing the vcore higher than 1.1V as it would die. Quote: "We have very fragile Dr.Mos chips and not very efficient in energy conversion, and mini-inducers, which anyway would be made redundant more than 30A mosfet, a combination very-very poor, I never thought I'll ever see it on a motherboard of this caliber." Essentially he had to over-cool it to run it beyond its specs and says its unlikely to survive.

Edit: In contrast, the 6990: http://lab501.ro/placi-video/asus-radeon-hd-6990-studiu-de-overclocking/10
4x 80A Volterra VRMs each GPU, or 8 x 80A = Well beyond anything you will need and you won't have to ever worry about it. The gtx580 OC to ~1ghz with his setup, pretty high OCs, the GPUs are capable.

Edit2: 675mhz under reference setup. http://lab501.ro/placi-video/asus-geforce-gtx-590-vs-hd-6990-clash-of-the-titans/13

 

[Idontcare]

The article is suggesting it does that under normal gaming loads. No other GPU or CPU I've ever heard of does that.

If true, this was planned. Get the reviewers to review at an unsustainable power draw then quietly downclock in the next driver release.

Unfortunately, thanks to the shenanigans of OCZ of late in regards to their SSD's, I can't argue against your position that Nvidia has indeed orchastrated an intentional "bait and switch" with the reviewers versus the end-suer

We have a real-world example of this sort of thing happening.

However, I would still argue that until we have proof Nvidia planned this change-up before they sent cards to the reviewers then it is far more likely that there is nothing more sinister afoot here than a dejavu P3 1.13GHz style of a cock-up.

 

[SlowSpyder]

Eh, you can never tell on the internet. But I see why you would think it was a joke, it does sound like it would be sarcasm.

I assure you I do not have a tungsten case. I thought suggesting I melted through it with a 4MHz core while playing Minesweeper would be dripping with enough sarcasm. But, I added that note to the post... apologies for any confusion.

 

[Soleron]

(removed)

 

[Tempered81]

The reviewer specifically said the 590 should have had 6 Volterra 40A VRMs per core rather than the 5x bad TDA 35A with low efficiency. It limited him pushing the vcore higher than 1.1V as it would die. Quote: "We have very fragile Dr.Mos chips and not very efficient in energy conversion, and mini-inducers, which anyway would be made redundant more than 30A mosfet, a combination very-very poor, I never thought I'll ever see it on a motherboard of this caliber." Essentially he had to over-cool it to run it beyond its specs and says its unlikely to survive.

Edit: In contrast, the 6990: http://lab501.ro/placi-video/asus-radeon-hd-6990-studiu-de-overclocking/10
4x 80A Volterra VRMs each GPU, or 8 x 80A = Well beyond anything you will need and you won't have to ever worry about it. The gtx580 OC to ~1ghz with his setup, pretty high OCs, the GPUs are capable.

Yeah he says the voltage circuitry on the 6990 is more suitable for Welding rather than rendering Graphics.

AMD are providing each Cayman GPU with 4x80AMP inductors, and Nvidia gives each GF110 GPU 5x35AMP inductors.

 

[notty22]

The 6990 also throttled during testing here at Anands.
Everything people are panicking about is also part of AMD's power strategy with the 6990.
http://www.anandtech.com/show/4209/amds-radeon-hd-6990-the-new-single-card-king/3

By playing with power management we’re of course referring to PowerTune, which was first introduced on the 6900 series back in December. By capping the power consumption of a card at a set value and throttling the card back if it exceeds it, AMD can increase GPU clocks without having to base their final clocks around the power consumption of outliers like FurMark. The hardest part of course is finding balance – set your clocks too high for a specific wattage and everything throttles which is counterproductive and leads to inconsistent performance, but if clocks are too low you’re losing out on potential performance.

This is without overclocking at 830mhz default bios

 

[SlowSpyder]

The 6990 also throttled during testing here at Anands.
Everything people are panicking about is also part of AMD's power strategy with the 6990.
http://www.anandtech.com/show/4209/amds-radeon-hd-6990-the-new-single-card-king/3

There is one notable difference between the 6990's and GTX590's though. That is, to my knowledge there haven't been multiple reports (some in product reviews) of the 6990 going up in smoke. They're also not on their fourth driver in five days.

 

[notty22]

There is one notable difference between the 6990's and GTX590's though. That is, to my knowledge there haven't been multiple reports (some in product reviews) of the 6990 going up in smoke. They're also not on their fourth driver in five days.

WTH are you responding to ? My post is pointing out power strategies both AMD and Nvidia seem to be using with these dual near 300 watt cards. Is that going to be the standard bs response to every angle of discussion from now on ?
But but but they blow up !

 

[jimbo75]

No the main difference is, the 6990 didn't get reviewed and benchmarked with a relaxed throttling unlike the 590 did.

 

[Jionix]

There is one notable difference between the 6990's and GTX590's though. That is, to my knowledge there haven't been multiple reports (some in product reviews) of the 6990 going up in smoke. They're also not on their fourth driver in five days.

Nor has AMD released multiple drivers that have lowered the power threshold each time, thus lower performance.

As far as I have heard, AMD set one power threshold with the cards release, and hasn't altered it.

 

[WelshBloke]

WTH are you responding to ? My post is pointing out power strategies both AMD and Nvidia seem to be using with these dual near 300 watt cards. Is that going to be the standard bs response to every angle of discussion from now on ?
But but but they blow up !

Well the whole point of those power strategies is to avoid the cards blowing up, so I cant see how its bad too point out that one of the strategies isn't working.

 

[Grooveriding]

So with the latest driver the GTX 590 is downclocking below the 600Mhz mark in demanding gaming conditions to prevent exploding.

Hopefully we see reviewers revisit the card with the new drivers now, to explain what is happening and the reduction in performance.

If it was me, I would be returning the card for a full refund. Not just due to the fact they are faulty and go up in smoke, but for drivers reducing performance from what is advertised on the box and on nv's website.

Turd indeed.

 

[pcm81]

The 6990 also throttled during testing here at Anands.
Everything people are panicking about is also part of AMD's power strategy with the 6990.
http://www.anandtech.com/show/4209/amds-radeon-hd-6990-the-new-single-card-king/3

This is without overclocking at 830mhz default bios

6990 does throttle if it gets too hot. My 6990 throttles down to as low as 550MHz and drops VCore to reduce temperature of the GPU. I've seen it as low as 550MHz, but I have 2 cards side by side. Water cooling should solve this issue.