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

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

tigersty1e

Golden Member
Dec 13, 2004
1,963
0
76
these dual gpu cards are just a drain on resources that the companies are only bringing to market for bragging rights.

the whole market would be done more justice if they did the same binning on these cards, but released superclocked versions of the gtx 580/6990.
 

Morg.

Senior member
Mar 18, 2011
242
0
0
On the other hand, expecting two 300Watts to sit on a 375Watt board and nothing bad happening sounded quite realistic.

And in the opposite corner, the 6990, sporting two 200Watt Gpus on a 375Watt board ... does indeed look less dumb... for no apparent reason.

Even though the nVidia cards should not explode, I have trouble conceiving that one could cool two 580's at full speed (600watts) on air, without a noise level much north of 80dB.

Now take that to overclocking levels and you have to take out 800+ watts with air from a pcb thats 11" ... right. You see that towercooler on your CPU ? it takes out about 200watts if you got a hard OC . Now just attach 4 of these on your graphics board and you'll be fine, and quiet. (yes your graphics board now weighs more than 6 lb's and your pcie slot is dying from the pressure and there's not enough room in your case, and goddamit stop whining just watercool it)

AS a summary, such a graphics board without watercooling makes no sense at all, and I sure hope the gods of overclocking smite down the fools who made those aircooled reference designs.
 

taltamir

Lifer
Mar 21, 2004
13,576
6
76
Even though the nVidia cards should not explode, I have trouble conceiving that one could cool two 580's at full speed (600watts) on air, without a noise level much north of 80dB.

They are not full speed 580s. They are have had their clockspeeds massively reduced.
 

PingviN

Golden Member
Nov 3, 2009
1,848
13
81
40nm made a run for it, broke free and made a run for it. Almost made it before heat and power took it down. These products really should've been on 32nm. Or not existed at all. Both HD6990 and GTX590 are... off.
 
Feb 19, 2009
10,457
10
76
40nm made a run for it, broke free and made a run for it. Almost made it before heat and power took it down. These products really should've been on 32nm. Or not existed at all. Both HD6990 and GTX590 are... off.

Disagree. People who use CF/SLI 6950/gtx570 or above deal with the same issues, except regular CF/SLI has much higher power use.
 

PingviN

Golden Member
Nov 3, 2009
1,848
13
81
Disagree. People who use CF/SLI 6950/gtx570 or above deal with the same issues, except regular CF/SLI has much higher power use.

They also have a much easier time cooling the cards down. I'd pick Crossfire HD6950/6970 over HD6990 any day.
 
Feb 19, 2009
10,457
10
76
Essentially the difference between a 2x6970 and 1x6990 is that you have to manually OC the 6990 to 880 core clock, without a voltage bump it doesnt add much extra power so its still a sizeable difference in power use. Its running within spec so it wont die from overheating. The only issue is noise. If you build your case well and place it 1m away from your head when you game, 47 dB is no biggie.

Personally, i would get CF 6950 with a bios mod/OC and call it a day. Quiet, efficient and great performance. A lot cheaper too.
 

Will Robinson

Golden Member
Dec 19, 2009
1,408
0
0
40nm made a run for it, broke free and made a run for it. Almost made it before heat and power took it down. These products really should've been on 32nm. Or not existed at all. Both HD6990 and GTX590 are... off.

I wonder if we'll just see the manufacturers turn up the memory and engine clocks of the 28mm cards until they reach 'typical operating temperatures"?
Even with the dieshrink and architecture changes I'd wager we won't see their high performance cards reach the H/P/N numbers we are all hoping for.
 

PingviN

Golden Member
Nov 3, 2009
1,848
13
81
Essentially the difference between a 2x6970 and 1x6990 is that you have to manually OC the 6990 to 880 core clock, without a voltage bump it doesnt add much extra power so its still a sizeable difference in power use. Its running within spec so it wont die from overheating. The only issue is noise. If you build your case well and place it 1m away from your head when you game, 47 dB is no biggie.

In my opinion, the difference is that instead of 2 coolers, you get 1. And HD6990 is far from quiet.

Personally, i would get CF 6950 with a bios mod/OC and call it a day. Quiet, efficient and great performance. A lot cheaper too.

^Of this I approve.
 

imaheadcase

Diamond Member
May 9, 2005
3,850
7
76
OCP fail. A transistor is the fastest fuse known to man. (or woman!)

Too bad such a catastrophic failure makes an expensive paperweight!

I'm surprised this happens with this design. But it did. Two years and back to the drawing board. D:


How are you surprised by this? Nvidia said they have to use binned chips just to get that performance they did. Does anyone really surprised overclocking it would cause it to brick..

I swear people here just think making a dual chip card is slapping to pieces of silicon together..
 

sandorski

No Lifer
Oct 10, 1999
70,128
5,657
126
So many miscues by Nvidia the last couple years. This wouldn't have happened if the Drivers were better Tested/Manufacturers were told not to Ship with them. All this will catch up with them.
 

Morg.

Senior member
Mar 18, 2011
242
0
0
They are not full speed 580s. They are have had their clockspeeds massively reduced.

I mean :
If you want this board to be worth anything, you want it full speed, and you're going to clock it that way
And, in that case blablabla
 

TerabyteX

Banned
Mar 14, 2011
92
1
0
Guys I remembered a while a ago either faulty drivers or hardware causes nvidia cards to self destruct when playing star craft II - it seems this time it went up another notch!

The next set of drivers may melt your videocard ala Chernobyl and kill with radiation poisoning!!!
 

Seero

Golden Member
Nov 4, 2009
1,456
0
0
If video card/gpu makers isn't going in the wrong direction, we can soon needs to replace our 20A circuit breaker with 40A and only connects to 1 PC and not anything else; need a multi-phase cooler to keep it from exploding; and a standalone electricity generator.

I guess, it will make more sense to simply have it run on gas.
 

Mopetar

Diamond Member
Jan 31, 2011
8,016
6,466
136
Guys I remembered a while a ago either faulty drivers or hardware causes nvidia cards to self destruct when playing star craft II - it seems this time it went up another notch!

I think it was some code in SCII that didn't cap the frame rate for some menu which would cause the card to run at full load and could lead to excessive wear on the cards causing some to fail. Unless there was some other problem with SCII that I don't remember hearing about.
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
How are you surprised by this? Nvidia said they have to use binned chips just to get that performance they did. Does anyone really surprised overclocking it would cause it to brick..

I swear people here just think making a dual chip card is slapping to pieces of silicon together..

That's not the point. I'm willing to bet the RAM and GPUs are fine. This is a power delivery issue most likely. There should be safeguards in hardware that prevent VRM components from approaching and ultimately exceeding SOA.

Nowhere did I suggest "making a dual chips card is slapping to pieces of silicon together" [sic]
 

Madcatatlas

Golden Member
Feb 22, 2010
1,155
0
0
So is this why the reviewers asked for more time?


you think so? you know Keysplayr mentioned they needed more time...mail time maybe?


lol

--------------------------------------

The inline image is unacceptable for this forum.

Take some time off to read the forum guidelines

Common courtesy
AT Admin
 
Last edited by a moderator:

chewietobbacca

Senior member
Jun 10, 2007
291
0
0
5970 was great, they were made to OC and most managed to reach and exceed 5870 speeds. no problems with overheating or dying cards.
Again, 6990 are the same, most are able to be manually OC beyond 6970 speeds and dont die from overheating or component failures. Perhaps there is a reason why these cards and their PCBs are so long and robust.

Whats worrying about the 590s is the crazy temps on the actual card/PCB itself, reaching a staggering 112C under gaming load. That puts out a lot of radiant heat to nearby hardware. SLI 590s will likely die very fast unless its on water.

If these cards keep on dying even with the new fixed drivers, it will be a huge problem.

The 6990 feels like it was designed for its purpose. The 590 feels like it got slapped together at the last second
 

Qianglong

Senior member
Jan 29, 2006
937
0
0
I thought those modern GPU boards have bios/hardware safety shut off feature when the boards get to a certain temp?
 

WelshBloke

Lifer
Jan 12, 2005
30,986
8,699
136
I thought those modern GPU boards have bios/hardware safety shut off feature when the boards get to a certain temp?

Nvidias is still (partly?) implemented in software.

Edit: And it looks like its the voltage circuitry thats cr*ping out and not the GPU's failing due to heat.
 
Last edited:

ArchAngel777

Diamond Member
Dec 24, 2000
5,223
61
91
you think so? you know Keysplayr mentioned they needed more time...mail time maybe?


lol

--------------------------------------
Mod callouts are unacceptable

The inline image is also unacceptable for this forum.

Take some time off to read the forum guidelines

Common courtesy
AT Admin

Was the mod callout removed from his post? Because the post above the way it is currently is not a mod callout (at least, I don't see how it is). Anyway, the post above is not safe for work and should be removed (by you hopefully). I clicked on it and said "Crap!" and closed out of it immediately. Not cool. If you are wondering why I clicked on it - because I thought the bolded text was his signature (I don't read them) and when I went to quote his post to say NSFW, I found the mod comment... Just as a FYI... The whole ---------------- makes it look like a signature. Secondly, if the picture was not appropriate, I would have expected the link severed or removed. Anyway, just giving a regular forum members perspective.


Per AnandTech Forum Guidelines, all inquiries into moderator actions are to be strictly handled through the Moderator Discussions sub-forum.

Moderator challenges are not acceptable as they iniate thread-derail.

Idontcare
Super Moderator
 
Last edited by a moderator:

bryanW1995

Lifer
May 22, 2007
11,144
32
91
it was combo of way too much much voltage and a beta driver not intended for use on this card.

“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. Rest assured that GTX 590 operates reliably at default voltages, and our 267.84 launch drivers provide even more additional levels of protection for overclockers. For more information on overclocking and overcurrent protection on GTX 590 please see our knowledge base article here: http://nvidia.custhelp.com/cgi-bin/n...p_faqid=2947.”

interesting story, keys got an email from nvidia explaining that they had told the reviewers to only use the most updated drivers. However, Apoppin at alienbabeltech got his card straight from nvidia and they never told him to not use the drivers on the cd.
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
if Nvidia did not want reviewers to overclock the GTX 590, then I'm sure they would have stated so when sending them out

seems like Nvidia rushed and didn't internal test their cards well enough to foresee something like this happening

No, they want reviewers to use 30% factory overclocked cards, failing that the reviewers are supposed to OC the cards themselves...


5970 was great, they were made to OC and most managed to reach and exceed 5870 speeds. no problems with overheating or dying cards.
Again, 6990 are the same, most are able to be manually OC beyond 6970 speeds and dont die from overheating or component failures. Perhaps there is a reason why these cards and their PCBs are so long and robust.

Whats worrying about the 590s is the crazy temps on the actual card/PCB itself, reaching a staggering 112C under gaming load. That puts out a lot of radiant heat to nearby hardware. SLI 590s will likely die very fast unless its on water.

If these cards keep on dying even with the new fixed drivers, it will be a huge problem.

huh? 6990 is way too long, and the fan is louder than my 5 year screaming at the top of her lungs (almost). The only reason the 6990 looks decent right now is in comparison with the failure(s) of gtx 590. 5970 looked so good b/c it was out several months before anything could get anywhere near it, and when NV finally slapped together a single gpu card that got close it still generally lost by ~ 20% and was louder/hotter/more power hungry than the dual gpu 5970. Basically 5970 had no competition, AMD knew by the time it came out that it wouldn't have competition, and they were able to be relatively conservative with it. 6990 was pushed pretty hard by NV, in fact the ultra robust design and afterburner blower motor on it are a direct result of AMD's determination to keep it on top at all costs. Anybody looking at anything less than 3 gpus from either NV or AMD would be very foolish to grab one of these sandwiches.
 
Last edited:

ArchAngel777

Diamond Member
Dec 24, 2000
5,223
61
91
No, they want reviewers to use 30% factory overclocked cards, failing that the reviewers are supposed to OC the cards themselves...

I was impressed with how well the 6990 did. Performance per watt definitely favors AMD, which makes them more ideal on notebooks. I am still not a huge fan of the 6950 and 6970, but the 6990 is pretty impressive. Makes me think that if AMD wanted to make a monolithic chip, they could have made something quite a bit faster than the current 6970. Heck, even if they double Cypress it would still be about the size of of the 580 GTX, while probably outperforming it by 30%. Although, that is speculation on my part. There are obviously huge hurdles when dealing with increasing chip sizes.
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
So many miscues by Nvidia the last couple years. This wouldn't have happened if the Drivers were better Tested/Manufacturers were told not to Ship with them. All this will catch up with them.

Just like a sports team, they are on cruise control when way in the lead but when pushed they'll sometimes make mistakes. NV had a couple of years where ati was clearly inferior and NV could do no wrong. At least for now, those days are over, and you'll likely see more miscues from them (and AMD) in the future if the competition remains this tight.

re: madcatatlas' post: absolutely inappropriate picture, but I admit that I laughed nontheless.
 
Last edited:
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |