Overclock.net banner

Status
Not open for further replies.
1 - 20 of 1259 Posts

·
Premium Member
Joined
·
21,169 Posts
Discussion Starter #1
In light of the recent deaths of 570's, I'd like to gather some information to keep others from killing theirs!

Please fill out the following if yours died. Also please include a picture if you have one:

User:
Core Voltage:
Clocks:
OCP/Power limit disabled: (Y/N)
Load Temperature:
VRM Cooling:
What you were doing: (Gaming, benching, stress testing)

6d172f2f.jpg


If you'd still like to purchase a GTX 570, I urge you to get one of the following:
Quote:
Originally Posted by bleda76
-Gainward 570 Phantom, Gainward 570 GS Goes Like Hell, Palit 570 Sonic Platinum all have 8 phase VRMs (6 GPU and 2 memory). Same amount with the GTX 580. All these cards have identical PCB since they belong to the same graphics card manufacturer group
http://images.bit-tech.net/content_i...phantom-1b.jpg

- Asus GTX570 DirectCU II also has 8 phase VRMs (6 GPU and 2 memory)
http://www.techpowerup.com/reviews/A...ages/front.jpg
See these threads for more information:

http://www.overclock.net/nvidia/928291-disabled-power-limit-occt-fried-gtx-2.html#post12209267
http://www.xtremesystems.org/forums/showthread.php?p=4727355#post4727355
http://www.overclock.net/nvidia/914105-gtx-570-dead.html


User: xxbassplayerxx
Core Voltage: 1.212V (1.223 with DMM)
Clocks: 1100/1200
OCP/Power limit disabled: Nope!
Load Temperature: -40C
VRM Cooling: Stock unisink / cold from phase / fan
What you were doing: Running Vantage Performance

User: jonjryjo
Core Voltage: 1.1v
Clocks:900/1800/2100
OCP/Power limit disabled: Yes via /GTX500OCP gpu-z argument.
Load Temperature: 58*C, it never went above 45 with power limit enabled.
VRM Cooling: DDGTX580
What you were doing: OCCT

User: slickric21
Core Voltage: 1.05
Clocks: 880/2000
OCP/Power limit disabled: Yes, via GPUZ 1.5.1 & /GTX500OCP
Load Temperature: 44'C
VRM Cooling: Accelero Xtreme VR sinks (+2 extra 120mm fans blowing on them)
What you were doing: OCCT GPU, full screen/complexity 8.

User: kerrk
Core Voltage: 1.2v
Clocks: 920/1840/2200
OCP/Power limit disabled: No
Load Temperature: 70 C
VRM Cooling: Stock
What you were doing: Benchmarking with 3dMark 11

User: rancor
Core Voltage:1.100(After burner)1.034 GPU-Z
Clocks:875/1750/2000
OCP/Power limit disabled: Yes (Hit 105A in GPU-Z)
Load Temperature: 80
VRM Cooling: Stock
What you were doing: Furmark stress testing

User: mistkron
Core Voltage: 1.100
Clocks: 900/1800/2000
OCP/Power limit disabled: No!
Load Temperature:
VRM Cooling: Twin Frozr II Unisink
What you were doing: Unigine Heaven

User: Shredicus
Core Voltage: Unlocked to 1.2v was around 1.17 at the time
Clocks: 980/2200 - 920/2000
OCP/Power limit disabled: N
Load Temperature: 70c
VRM Cooling: Stock
What you were doing: Crysis

User: miahallen
Core Voltage: 1.2V via MSI AB v
Clocks: 950/2200
OCP/Power limit disabled: No
Load Temperature: ??? (fan @ 85%)
VRM Cooling: Stock
What you were doing: 3DMark Vantage

User: Pr0paIn
Core Voltage: 1025 mv
Clocks: 870 Core / 2075 Memory
OCP/Power limit disabled: No
Load Temperature: 74-76 Celcius
VRM Cooling: Stock
What you were doing: CryEngine 3 Sandbox

User: Basti51
Core Voltage: 1050mV
Clocks: 900/1800/2000
OCP/Power limit disabled: No
Load Temperature: 60°C
VRM Cooling: AXP
What you were doing: Gaming - Mafia 2

User: kenan435
Core Voltage: 1100mV
Clocks: 920/1840/2000
OCP/Power limit disabled: No
Load Temperature: 76°C
VRM Cooling: Stock
What you were doing: Furmark

User: Scorpion667
Core Voltage:1.1v
Clocks: 920/2000
OCP/Power limit disabled: never
Load Temperature: 76c
VRM Cooling: stock
What you were doing: haven 2.1

User: Riou
Core Voltage: stock
Clocks: stock
OCP/Power limit disabled: N
Load Temperature: 72C (Mirror's Edge)
VRM Cooling: stock
What you were doing: Browsing Overclock.net

User: whipgeez
Core Voltage: 1.038
Clocks: 850/2000
OCP/Power limit disabled: N
Load Temperature: 77c
VRM Cooling: Stock
What you were doing: Furmark

User: McDown
Core Voltage: 1.1v
Clocks: core 935, mem 2000
OCP/Power limit disabled: N
Load Temperature: 80C
VRM Cooling: stock
What you were doing: (Crysis 2, 270.51 Beta driver)

User: Officer Farva
Core Voltage:1.1v
Clocks: 900/2000
OCP/Power limit disabled: Nope
Load Temperature: 66
VRM Cooling: stock
What you were doing: Crysis (So no, it couldn't run it )

User: Chris Ihao
Core Voltage: 1.1V
Clocks: 890/1000
OCP/Power limit disabled: No.
Load Temperature: sub 80c
VRM Cooling: AC Xtreme Plus (VR004 sinks)
What you were doing: Shogun 2 with depth of field enabled. Wicked hot.

User: AzN1337c0d3r
Core Voltage: 1.05V
Clocks: 825/2200
OCP/Power limit disabled: (Y/N) N
Load Temperature: ~85
VRM Cooling: Stock
What you were doing: Furmark, boom after 20 seconds. Rebooted to Windows fine, would power off system when trying to launch 3D game.

User: AzN1337c0d3r
Core Voltage: Stock
Clocks: 850/2350
OCP/Power limit disabled: (Y/N) N
Load Temperature: ~85
VRM Cooling: Stock
What you were doing: Furmark, boom after 5 seconds. Would not POST after.

User: Born2bwild
Core Voltage: 1.100 V
Clocks: 920Mhz/2200Mhz
OCP/Power limit disabled: No
Load Temperature: 82*C
VRM Cooling: Stock, plus two 2000rpm 120mm fans blowing on the SLI stock
What you were doing: Crysis 2

User: bleda76
Card: Gigabyte GTX 570 OC (GV-N570OC-13I)
Core Voltage: 1.00V (factory default)
Clocks: 850Mhz core /2250Mhz memory
OCP/Power limit disabled: No
Load Temperature: 65-67C
VRM Cooling: Gigabyte Windforce x3
What you were doing: Furmark (3 minutes)

User: El_Capitan
Core Voltage: Stock 963mV
Clocks: Stock 732MHz
OCP/Power limit disabled: No
Load Temperature: Unknown
VRM Cooling: Stock
What you were doing: Playing StarCraft II
 

·
Registered
Joined
·
433 Posts
Good post.

Core Voltage: 1.1v
Clocks:900/1800/2100
OCP/Power limit disabled: Yes via /GTX500OCP gpu-z argument.
Load Temperature: 58*C, it never went above 45 with power limit enabled.
VRM Cooling: DDGTX580
What you were doing: OCCT

Will post pictures later.
 

·
Premium Member
Joined
·
6,171 Posts
Crap. I was going to do this cause I thought it might help me get a stable 900mhz overclock @ 1.1v.

Sorry for your loss, but thank God I saw this before I ended up turning off OCP.
 

·
Registered
Joined
·
44 Posts
Core Voltage: 1.2v
Clocks: 920/1840/2200
OCP/Power limit disabled: No
Load Temperature: 70 C
VRM Cooling: Stock
What you were doing: Benchmarking with 3dMark 11
 

·
I.T. Specialist III
Joined
·
4,670 Posts
Too many people are trying to hit high overclocks without proper knowledge, tuning and settings.

Because its so easy to turn it up on the GF-110's, I see many people with average case cooling and basic overclock knowledge hurting their GPU's. 900 core is my absolute max, after that my memory chips are nearly on fire. And thats with 65% fan speed during load.
To run these GPU's at 900+ core clocks or near there, you need more then sufficient cooling and tweaks to your fan speeds during load.

You also need crazy airflow.
It took me nearly a week of testing my GPU with tests and burns while watching temps and tuning the results to get it to where it is now. Just because your temps aren't high doesn't mean its stable. You need to look carefully at the benchmarks results to see where your FPS drops, and other performance related measuring tables.

You don't just dial it up and turn the fans up higher...basing your overclock on luck/assumptions is a road to RMA.
 
  • Rep+
Reactions: jcrandall623

·
Premium Member
Joined
·
21,169 Posts
Discussion Starter #7
The problem with the 570's is that they throttle to keep power within limits. Because of this maximum power draw limit, they were not designed to handle more (as most cards are). Even if you increase cooling substantially (my core was at -40°C under load, VRM's were definitely around 0°C), it cannot handle the increased load because it lacks the robust power delivery needed.

The goal of this thread is to collect information to pinpoint the issue so no one else experiences this.
 

·
Registered
Joined
·
433 Posts
Quote:
Originally Posted by amstech;12211658
Too many people are trying to hit high overclocks without proper knowledge, tuning and settings.

Because its so easy to turn it up on the GF-110's, I see many people with average case cooling and basic overclock knowledge hurting their GPU's. 900 core is my absolute max, after that my memory chips are nearly on fire. And thats with 65% fan speed during load.
To run these GPU's at 900+ core clocks or near there, you need more then sufficient cooling and tweaks to your fan speeds during load.

You also need crazy airflow.
It took me nearly a week of testing my GPU with tests and burns while watching temps and tuning the results to get it to where it is now. Just because your temps aren't high doesn't mean its stable. You need to look carefully at the benchmarks results to see where your FPS drops, and other performance related measuring tables.

You don't just dial it up and turn the fans up higher...basing your overclock on luck/assumptions is a road to RMA.
Although what you said is true, these failures have nothing to do with that. It's the simple fact that the 570 VRMs can not take the voltage with the power limiter unlocked. Simply overclocking won't kill a card unless you run it way over spec long term, although it will artifact and bsod long before that, which should be a telltale sign of instability. These VRMs simply destroy themselves without warning, no artifacts, no BSOD, the computer simply shuts off, and the card is done for.

I have a waterblock, so obviously heat was not an issue, and the OP was running sub-zero temps, the VRM's simply can not take the voltage required to run the card when the power limiter is disable.
 

·
Premium Member
Joined
·
21,169 Posts
Discussion Starter #9
It seems that might not be the only thing causing it.

Over at XS, W1zzard killed his just by overvolting. OCP still in place:
Quote:
Originally Posted by W1zzard @ XS
i blew a gtx 570 just by overvolting and stress testing with ovp,ocp,nvidia power limit still in place
http://www.xtremesystems.org/forums/showpost.php?p=4727640&postcount=19

The thing we're not used to is the fact that the voltage isn't killing the GPU, it's killing the power delivery system... and at much lower voltages than one would think.
 

·
Registered
Joined
·
711 Posts
what about the 580 does its system have simular problems ?

that really anoying

making me want to order 2 GTX480's and dont think on all that problems... or move to AMD side :\ 6970's seem like a good opinion
 

·
Premium Member
Joined
·
4,635 Posts
What about GTX480 are there any issues similar to GTX570 in any way?

I have tested my GTX480 with Furmark 20min Extreme mode @ 850/1025 and 1075mV set in Afterburner and it has pssed it with flying colors but themps have reached 100C(max) on the core so I have opened window to drop it to 90C..

Are there any deaths of GTX480's with blown VRM's doe OCing like on GTX570?

CHEERS..
 

·
Premium Member
Joined
·
21,169 Posts
Discussion Starter #17
A GTX 580 will not have these issues from standard use. It has 50% more phases than the 570 (six instead of four).
 

·
Registered
Joined
·
315 Posts
Quote:
Originally Posted by KingT;12214452
What about GTX480 are there any issues similar to GTX570 in any way?

I have tested my GTX480 with Furmark 20min Extreme mode @ 850/1025 and 1075mV set in Afterburner and it has pssed it with flying colors but themps have reached 100C(max) on the core so I have opened window to drop it to 90C..

Are there any deaths of GTX480's with blown VRM's doe OCing like on GTX570?

CHEERS..
I blew the VRM on my GTX 480:wave2:
Glad I bought the Zotac AMP edition after that. Runs cooler and dont have to be worried about noise and my VRMs anymore.

R.I.P EVGA GTX 480:sad-smile
 
1 - 20 of 1259 Posts
Status
Not open for further replies.
Top