Sign in with
Sign up | Sign in
Your question

Overclocking Instability Generic Questions (Two Scenarios)

Tags:
  • Graphics Cards
  • Overclocking
  • Core
  • Blue Screen
Last response: in Overclocking
Share
January 25, 2013 10:44:42 AM

Hi all,

I have a couple of general questions about overclocking instability. My apologies if this has been answered a million times before but there were no stickies.

Basically what are the most common symptoms of instability when O/C & running stress tests?

Scenario 1: Core clock isn't that high (compared to most overclocks on same card) but voltage is too high - What am I most likely to see? (temps increase to lets say mid 90's Celsius and then... (BSOD, black screen, temps keep increasing or what should happen next) )

Scenario 2: Reverse: Core clock is high but voltage is too low. Most likely to see artifacts/tearing/BSOD? Should your temps stay reasonably low (due to low voltage) or will the fact that you're stressing the card increase the temps naturally (again to around mid 90's)?

If anyone knows any other scenarios please let me know as well. At the moment I'm staying away from a memory O/C as I heard they can be very sensitive to O/C.

P.S. I have gtx 470 (i know its ancient) but the shader and core clock sliders are dependent in MSI AB. If I increase core clock, shader increase proportionally. Is that just a something in the setting because I have seen vids of people setting them independently.

Cheers all

More about : overclocking instability generic questions scenarios

January 25, 2013 11:17:17 AM

Scenario 2 is what I've experienced while testing. Raising the core clock without touching the voltage resulted in flickering windows and general inconsistency in the display - but the temperatures stayed roughly the same.

Bumping the voltage up a bit resolved it.
January 25, 2013 12:32:36 PM

Meoricin said:
Scenario 2 is what I've experienced while testing. Raising the core clock without touching the voltage resulted in flickering windows and general inconsistency in the display - but the temperatures stayed roughly the same.

Bumping the voltage up a bit resolved it.


Cheers for your reply. I'm a little confused by your answer though. What were you doing when there was flickering?

Do you mean there was flickering while just working away in windows (OS) and opening up and moving window panes (i.e. gpu not under load) or the burn in test was flickering (gpu at full load)? If it was the burn in test, was it the furry donut or the background image that was flickering?


Related resources
Can't find your answer ? Ask !
January 25, 2013 12:57:03 PM

I have also noticed (at least I think do) that the background image in Furmark (the tunnel effect) can have the tiniest little flicker (i.e. it pauses for 0.001 of a second and continues as normal). It is extremely hard to see and I am not sure if I'm even seeing it, its that tiny. Is this something that is just in the program or is it a sign of instability?
January 25, 2013 3:39:19 PM

smjpl said:
Cheers for your reply. I'm a little confused by your answer though. What were you doing when there was flickering?

Do you mean there was flickering while just working away in windows (OS) and opening up and moving window panes (i.e. gpu not under load) or the burn in test was flickering (gpu at full load)? If it was the burn in test, was it the furry donut or the background image that was flickering?


The flickering occurred while on the desktop, with virtually nothing open (Afterburner, Speedfan, Rainmeter & Foobar were the only windows).

Basically they'd flash in and out, with flickers of black/white across the desktop. Was quite alarming when it happened at first. I could use Afterburner to turn the core clock up to the standard levels fine (840mhz is what it limited it to), but once I unlocked that and turned it up more, the flickering would start almost immediately. 850/860mhz only had an issue when interacting with windows, but was fine if I did nothing - anything above that resulted in constant chaos.

Unlocked voltage control, bumped it up a notch or two, and have been running 900mhz since then. Going to turn that up more later.
January 25, 2013 4:02:02 PM

Meoricin said:
The flickering occurred while on the desktop, with virtually nothing open (Afterburner, Speedfan, Rainmeter & Foobar were the only windows).

Basically they'd flash in and out, with flickers of black/white across the desktop. Was quite alarming when it happened at first. I could use Afterburner to turn the core clock up to the standard levels fine (840mhz is what it limited it to), but once I unlocked that and turned it up more, the flickering would start almost immediately. 850/860mhz only had an issue when interacting with windows, but was fine if I did nothing - anything above that resulted in constant chaos.

Unlocked voltage control, bumped it up a notch or two, and have been running 900mhz since then. Going to turn that up more later.



Wow, I didn't think it would be that stressed under very little load. Did you do any stress tested on stock settings? Also what card is? Glad to hear all is well now though.
January 25, 2013 7:48:13 PM

It passes all the stress tests fine - no lock ups or crashes either. According to Afterburner it's always drawing the same power, so I figured it'd be that.

On a slight aside - earlier today I sleeved one of the PCI 6 Pin power leads (will do the other tomorrow), and after replacing it I had the flickering until I unplugged and adjusted the pins - suggesting that they weren't making a good connection, so it wasn't receiving the power it wanted.

I'm not convinced it's a matter of stress under load, simply the card needing more power than it's getting - whatever the reason for that is.

It's an XFX 1GB HD6950, with a Raystorm GPU waterblock on it.
!