Stable system not stable anymore?

dpk33

Senior member
Mar 6, 2011
687
0
76
I have a Phenom II 955 on a GIGABYTE MA790FXT-UD5P stable at 4 ghz at 1.55 volts for almost 3 years now (since the beginning of 2009). I gamed, I surfed the web, I rendered video. All rock solid. But out of curiousity, I decided to check my temps recently with Intel Burn Test. I got a blue screen. Now the blue screens have been becoming more and more frequent. I'm starting to get worried. I entered BIOS, upped the volts a bit, still unstable. Put it back to its old settings. Still unstable. What could be wrong?

EDIT: It's usually around 55 degrees under load before it blue screens.
 
Last edited:

Rifter

Lifer
Oct 9, 1999
11,522
751
126
not enough voltage, chips degrade over time and require more voltage to obtain the same clocks. So either up the voltage or lower clocks.
 

podspi

Golden Member
Jan 11, 2011
1,982
102
106
Yea, OP I'd run memtest overnight and see if that turns up anything before deciding it is the CPU.


Any info on what the bluescreen says? They can often be very useful at pinpointing what exactly is going wrong.

Edit: Also, I see that you ran IBT. Those types of stress programs do exactly that, and they can sometimes cause things to blow. Might also want to check your motherboard to see if a cap maybe couldn't take the pressure any more?
 

DominionSeraph

Diamond Member
Jul 22, 2009
8,386
32
91
"I put a hammer through my LCD. Now the picture doesn't look as good. What could be wrong?"

From experiance I would say your memory is going south.

So he should be able to get 300GHz stable if only his memory is good?
In all likelihood he fried his processor and now will be lucky to get stock speeds out of it.
 
Last edited:

taltamir

Lifer
Mar 21, 2004
13,576
6
76
Any time you run a current through your chip you are degrading it, physically. Overclocking makes it degrade faster, overvolting MUCH faster.

The main reason for so much overhead on the stock processor is because AMD needs it to work reliably for years. When overclocking and overvolting you sacrifice that longevity for short term performance gains.

3 years of overclocking/volting has degraded your processor to the point where it is no longer stable at stock speeds. Either replace it, or drop below stock speeds while pushing voltage even higher. This will allow you to get a little more work out of it but will significantly accelerate the rate at which your already damaged processor degrades, until it finally burns out... literally, as in catches on fire. So actually I recommend you don't do that and instead just buy a new processor.
 
Last edited:

VirtualLarry

No Lifer
Aug 25, 2001
56,587
10,225
126
to the point where it is no longer stable at stock speeds.
He didn't say that he dropped it to stock, just that he put it back to the "old settings", which I assumed meant his old overclock.

OP, I would drop things to stock, and then run a Memtest86+ overnight, as well as some OCCT, IntelBurnTest, or Prime95.
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
I have a Phenom II 955 on a GIGABYTE MA790FXT-UD5P stable at 4 ghz at 1.55 volts for almost 3 years now (since the beginning of 2009). I gamed, I surfed the web, I rendered video. All rock solid. But out of curiousity, I decided to check my temps recently with Intel Burn Test. I got a blue screen. Now the blue screens have been becoming more and more frequent. I'm starting to get worried. I entered BIOS, upped the volts a bit, still unstable. Put it back to its old settings. Still unstable. What could be wrong?

EDIT: It's usually around 55 degrees under load before it blue screens.

You just answered your own question. In case you didn't realise by now that is a LOT of voltage for 24/7.

Next time don't be so reckless, bruddah. 3.8GHz is easily doable with 1.35V, an extra 200MHz isn't worth the extra 0.2V.
 
Last edited:

taltamir

Lifer
Mar 21, 2004
13,576
6
76
He didn't say that he dropped it to stock, just that he put it back to the "old settings", which I assumed meant his old overclock.

Maybe I misread. Maybe you are assuming wrong.
If he merely rolled back to a lower OC then he should drop all the way to stock.
If he did drop it to stock and it is still not stable then we know why.
 

dpk33

Senior member
Mar 6, 2011
687
0
76
It's still oc'ed to 4ghz at 1.55 volts. i upped it to 1.575 but it did nothing. Still crashed. Then I put the voltage back to 1.55. What is a good voltage that I can set it at before i start to damage my chip?
 

taltamir

Lifer
Mar 21, 2004
13,576
6
76
It's still oc'ed to 4ghz at 1.55 volts. i upped it to 1.575 but it did nothing. Still crashed. Then I put the voltage back to 1.55. What is a good voltage that I can set it at before i start to damage my chip?

A chip is degraded if you run a current through it, period. Even heavily undervolting it, it still degrades. The more voltage you put through it, the faster it degrades. The increase is an exponential function, not linear. A heavily undervolted chip will last much much longer than a chip running at stock voltage.

Also, increasing the clockspeed ALSO increases degeneration speed, but not nearly as much as voltage increase.

Lower your voltage back to the chip default, then find what it tests stable at, then run it at 200-300 mhz below that.
 

tweakboy

Diamond Member
Jan 3, 2010
9,517
2
81
www.hammiestudios.com
Holy cow 1.55v is a lot . Im suprised that CPU didn't blow up. What type of cooling you have on it.

Depends, if the BSOD says, irq_not_less_or equal then its your RAM. Also try running the CPU at stock speed just for test to see if you get the BSOD, so we can rule out RAM possibly and its your CPU voltage too high over time possibility its lost its edge. BUt I would first put everything on stock speeds and see if you get the BSOD. Then well take it from there..
 

dpk33

Senior member
Mar 6, 2011
687
0
76
Holy cow 1.55v is a lot . Im suprised that CPU didn't blow up. What type of cooling you have on it.

Depends, if the BSOD says, irq_not_less_or equal then its your RAM. Also try running the CPU at stock speed just for test to see if you get the BSOD, so we can rule out RAM possibly and its your CPU voltage too high over time possibility its lost its edge. BUt I would first put everything on stock speeds and see if you get the BSOD. Then well take it from there..

For AMD cpus I don't believe that's a lot. It's +0.150 compared to the original. And yes I believe my blue screen did mention something like that. And what could have possibly happened to the memory to make it go bad?
 

aigomorla

CPU, Cases&Cooling Mod PC Gaming Mod Elite Member
Super Moderator
Sep 28, 2005
21,073
3,576
126
Im wondering how many people that said 1.5V is too much for an AMD actually owns an AMD?


LOLz...

Your head is stuck in the intel cloud my friend, and intel has its own set of rules.

From my knowledge 1.5V on an AMD processor is OK as long as u keep the heat in check.

For AMD cpus I don't believe that's a lot. It's +0.150 compared to the original. And yes I believe my blue screen did mention something like that. And what could have possibly happened to the memory to make it go bad?

Awwwww u had to make the intel people feel dumb now didnt ya.. lol...
Dont worry guys.. i used to be here as well... i was quickly edjucated..
Well more like face slapped with words lolol.....


BTW OP, i +1 Nemisis ... omg the sky has fallen...
From experiance I would say your memory is going south.
 
Last edited:

BD231

Lifer
Feb 26, 2001
10,568
138
106
1.5v on AMD chips is all bad, had a 955 for two years that I ran at 1.52 and the same thing happened.

AMD is full of it on the 1.5v+ spec.
 

TakeNoPrisoners

Platinum Member
Jun 3, 2011
2,599
1
81
That is quite a bit of voltage. I am suprised it lasted that long. If you want you can buy another X4 955 for about $100 now. I don't put my X4 955 over 1.4V and it is stable at 3.9GHz.

Most recommend not putting over 1.4V. You can run 1.45V if you are feeling brave.
 

geokilla

Platinum Member
Oct 14, 2006
2,012
3
81
My chip degraded a lot as well. Now I can barely push it past stock. Currently running it at 3.52Ghz with 1.475Vcore. Temps at 70C.

I blame the shitty Earthwatts for higher CPU load temps. Before it was 60C.
 

TakeNoPrisoners

Platinum Member
Jun 3, 2011
2,599
1
81
My chip degraded a lot as well. Now I can barely push it past stock. Currently running it at 3.52Ghz with 1.475Vcore. Temps at 70C.

I blame the shitty Earthwatts for higher CPU load temps. Before it was 60C.

At 70C your chip should have shut down. That is way above spec. The program must not be reading properly. Also clean dust out. It probably isn't your PSU.
 

geokilla

Platinum Member
Oct 14, 2006
2,012
3
81
At 70C your chip should have shut down. That is way above spec. The program must not be reading properly. Also clean dust out. It probably isn't your PSU.
It is my PSU. With the GameXStream 600W, I was doing 60C on both the CPU and GPU. After it died and I replaced it with the Earthwatts 500W, it's now 70C each. And yes I reseated the heatsink and dusted it out like I always do every couple months.
 

DominionSeraph

Diamond Member
Jul 22, 2009
8,386
32
91
It is my PSU.

No it's not. Wattage is voltage * current. For your CPU to be putting out more heat at the same load either the voltage needs to increase or the CPU's resistance needs to decrease. If your motherboard is supplying the same voltage then both PSU's input into this equation is the same.
Your heat issues are probably a combination of CPU degradation and heat sink misapplication. (Reseating every couple of months?? Really??)

Also, why the hell are you still pushing your CPU? 1.475V for a mere 10% overclock? You probably run that 24/7 too instead of letting Cool & Quiet downclock and downvolt, amirite? (I'm idling @ 1.075V on a 1.35V processor)

A destructive overclock on a budget processor is pretty stupid IMO. You end up spending more than it would've cost to just put in a faster processor to begin with. An i5 760 is as fast stock as a 955 is at 4GHz. You could've given one of those an easy 15% overclock and been good for years running cooler and faster than the 955.
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
A destructive overclock on a budget processor is pretty stupid IMO. You end up spending more than it would've cost to just put in a faster processor to begin with. An i5 760 is as fast stock as a 955 is at 4GHz. You could've given one of those an easy 15% overclock and been good for years running cooler and faster than the 955.

http://www.hardwaresecrets.com/article/819

According to that, i5 750 was $200 and 955 BE is $189 on Sep 09. To buy AMD at that price range, you have to be out of your mind when the 750 hits 3.6GHz at stock volts and has better IPC.
 

geokilla

Platinum Member
Oct 14, 2006
2,012
3
81
No it's not. Wattage is voltage * current. For your CPU to be putting out more heat at the same load either the voltage needs to increase or the CPU's resistance needs to decrease. If your motherboard is supplying the same voltage then both PSU's input into this equation is the same.
Your heat issues are probably a combination of CPU degradation and heat sink misapplication. (Reseating every couple of months?? Really??)

Also, why the hell are you still pushing your CPU? 1.475V for a mere 10% overclock? You probably run that 24/7 too instead of letting Cool & Quiet downclock and downvolt, amirite? (I'm idling @ 1.075V on a 1.35V processor)

A destructive overclock on a budget processor is pretty stupid IMO. You end up spending more than it would've cost to just put in a faster processor to begin with. An i5 760 is as fast stock as a 955 is at 4GHz. You could've given one of those an easy 15% overclock and been good for years running cooler and faster than the 955.
I dust it every couple months. I reseat it once a year.

I am a Folder. Does that explain things for you? And no I'm not gonna spend money on a new rig when this is perfectly fine. I'll use it till it blows up.

And if the PSU isn't causing ambient temperatures to go up, then please enlighten me. How is my GTX 460 folding at 70C with the Earthwatts 500W whereas I as at 60C with the GameXStream 600W? Simple. The PSU is working harder to supply the power my PC needs, the 80mm fan is an exhaust fan and not an 120mm intake fan like on all other PSUs. If it was just my CPU temps that increased and not the GPU, then we got a problem.
 

Nemesis 1

Lifer
Dec 30, 2006
11,366
2
0
Im wondering how many people that said 1.5V is too much for an AMD actually owns an AMD?


LOLz...

Your head is stuck in the intel cloud my friend, and intel has its own set of rules.

From my knowledge 1.5V on an AMD processor is OK as long as u keep the heat in check.



Awwwww u had to make the intel people feel dumb now didnt ya.. lol...
Dont worry guys.. i used to be here as well... i was quickly edjucated..
Well more like face slapped with words lolol.....


BTW OP, i +1 Nemisis ... omg the sky has fallen...

LOL I built many many AMDs lol . I just won't buy 1 for self . Someone wants one . They get one . LOL at me from you tho is hard to take . I recall THe bad things you had to say about SB befor it was released . Than that water cooled system you showed at XS with all the dust . I couldn't believe you would let a system get so dirty. So watch the lol there bud. LOL