• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

OC won't OC anymore -- UPDATE, SEE INSIDE, ITS BACK

Markfw

Markfw

Moderator Emeritus, Elite Member
I had a perfect OC at 3150, running 24/7 for 60 days or so, now today, it started rebooting every 60 seconds...No hardware or software changes at all.

So I take it down to 3060, and ir runs for 5 hours, and re-boots again.

Now down to 3045, and we will see how long that works for.

WTF ????

Edit: It is a B3 stepping.....Not over 70c for those 2 months using coretemp.
 
You know, I can't believe nobody cares about this or is curious. I have 4 quads, and this was the first. So this is what is to come for the rest ?
 
what cooling are you using on that unit? what vcore? don't you have 2 b3's? What are the build dates? Also, what are the other particulars of the unit, esp mobo/ram/case w airflow?
 
This is weird I have no clue what is going on. How much juice have you been giving it?

Edit:god bryan just beat me out and did a better job.
 
Processors can develop a 'tolerance' to increased voltages over time. I know a guy w/ an E4300 on water that used to run 3.6 24/7 and a year later can only run 3.5 24/7 because it needs more Vcore (1.6V presently). As an example from my experience:

This is my CPU: E4400 @ 300FSB, 10x for 3.0GHz @ 1.345Vcore (used 24/7, one step up from default Vcore)

This is my CPU on drugs: E4400 @ 340FSB, 10x for 3.4GHz @ 1.505Vcore (used to see what I can get for a stable overclock [using Orthos for stability/ Sandra for benching] and could run 24/7 if I had funds to get a new CPU once it developed the aforementioned tolerance.)

It's also a good idea to run the default multplier so you can keep C1E & EIST enabled, which will lower your voltage & multiplier while your CPU is idle, hence leading to a longer life.

If you've been running an increased voltage, try upping it by the smallest increment you can; it may just fix it.
 
mark is the folding monster god of AT, he runs his units at full load 24/7. I don't think that C1E and EIST are of much use to him.
 
Thanks for the replies ! Answers:

Cooling : Ninja Sythe, max temps 70c
Vcore 1.320 under loads (not sure bios, been a while, but has a lot of vdroop, probably 1.4675 as I remember.)
Case ANTEC 900 all fans high.
Motherboard is P5K (non deluxe)
Ram is Gskill PC 8000, way under spec at 435 currently, was 450, still 10% under spec. 2.2 vdimm.

Bottom line, is they don't like being OC'ed maybe.

I don't know the week, etc.., just B3 Stable for 5 hours, but at 3045. My other B3 is still at 3200, and it runs 72c, same cooling on an S3 mobo. My X3210 is still at 2800 (POS).

And my G0 stepping is still at 3.4 and cool, and fine.

Time will tell.
 
that's a beatiful example of the incredible difference you can get with the exact same or certainly very similar chips. 4 quads, max oc over 500mhz difference, no 2 of them are within 100 mhz (for now) of each other. Most are perfectly happy but one has developed "Folding24/7istoomuchforme-itis".

You obviously have great system parts. Bad luck? Bad batch?
 
Yup. I agree. My G0 has all the perfect parts. and will do 3.6 stable, but not F@H stable at that, just 3420.(I think the best test)

Mobo 😀Q6
Ram OCZ PC 6400 gold rev 2
PSU Seasonic S12 550
Cooler Ultra 120 lapped.

Edit: I don't get my B3 on an S3@3.2, its been that way since the days I installed it, my most stable (except the G0, since I set it to 3.4) But I won't mess with stability !
 
So right now vcore is an actual 1.320v at load? Do you remember if that's what it has always been when it ran stable before? Just trying to make sure that the cpu is still getting the same volts and it's not that the motherboard/psu are not giving the same volts they used to under load. Also you might want to try running memtest86 to eliminate ram as a problem and make sure it is indeed the cpu that's suddenly needing more volts to run stable. Only other thing I know that can effect a stable overclock is heat. If the temps have increased since when it was stable that can effect it, but summer is ending so I'd imagine your temps might have dropped if anything. If all that's checked then it sounds like the chip just didn't hold up at those settings for long.

On a side note my system is still a mobile Athlon XP barton running 2.4ghz @ 1.76vcore and it's still just as stable as when I built it 4 years ago. However a similar system I built for my cousin with an Athlon XP 2500+ barton lost it's stability after only a year but I've learned a lot since then and it could of been the ram/mobo/ect and not the cpu(I didn't check everything).
 
The HSF has no dust. Anyways, if it did, the temps would be higher. And the vcore has always been 1.320 under load, no changes in anything. (except the OC)
 
Originally posted by: Markfw900
You know, I can't believe nobody cares about this or is curious. I have 4 quads, and this was the first. So this is what is to come for the rest ?
Click to expand...

mark why dont you just PM me.

You know when it comes to overclocking and error debugging on these babies there are very few people that know more then i do.

What the problem seems to be is a board degrading issue. I cant stress this enough, on some boards the mosfet area cooling is poorly contacted. This leads to that portion of the board degrading. Also installing a sink with a poor backplate, or tightening that backplate to no other can lead to warpage issues. This will lead to poor contact on the mosfet area. I think you degraded a mosfet on your board.

It has happened on Abits, eVGA 680i and ASUS and lower end Gigabytes. Now your really going to thank me on why i recomended that DQ6 to you. That rear plate HELPS in preventing warpage.

Pull the sink off the mosfet area and check to see how many spots on the pad has indentations. IF you see all indentations on the pad, then somewhere else your board had problems.


http://www.xtremesystems.org/f...ghlight=mosfet+problem

Here is a simular situation to what your running. Try to secure the mosfet cooling area, or try to put some enzo ram sinks over them and see if that can save your board. Otherwise a RMA looks very possible in your near future.
 
In my experience when a previously well functioning system starts to shut down and reboot it is symptomatic of a failing PSU. If you have a spare try swapping it out and see if that doesn't fix things.

Oh, I don't have the god-like expertise of some but I have been doing this for about 14 years now. 🙂
 
Its a brand new (2-3 months old, whole system was new) 700 watt Fortron. I can't believe its overloaded and failing. I don't have anything better to swap, maybe a Antec 450.

aigomorla: Its a P5K ASUS (non-deluxe). With those two Antec 900 fans blowing right over the motherboard, one pulling, and the 200mm pulling the other direction, I have plenty of airflow. I may need some ramsinks over the MOSFETS.
 
Originally posted by: hokiealumnus
Processors can develop a 'tolerance' to increased voltages over time. I know a guy w/ an E4300 on water that used to run 3.6 24/7 and a year later can only run 3.5 24/7 because it needs more Vcore (1.6V presently). As an example from my experience:

This is my CPU: E4400 @ 300FSB, 10x for 3.0GHz @ 1.345Vcore (used 24/7, one step up from default Vcore)

This is my CPU on drugs: E4400 @ 340FSB, 10x for 3.4GHz @ 1.505Vcore (used to see what I can get for a stable overclock [using Orthos for stability/ Sandra for benching] and could run 24/7 if I had funds to get a new CPU once it developed the aforementioned tolerance.)

It's also a good idea to run the default multplier so you can keep C1E & EIST enabled, which will lower your voltage & multiplier while your CPU is idle, hence leading to a longer life.

If you've been running an increased voltage, try upping it by the smallest increment you can; it may just fix it.
Click to expand...

I think the "tolerance" you are talking about is something in EE that they call "hot electrons."

When you over-volt a CPU beyond the specs, the transistors in the chip experience high electric fields which cause, "hot electrons" or electrons that have extremely high energy. These high energy electrons can enter the oxide region in the gate where they can become trapped causing oxide charging. Over time, these charges will accumulate and degrade the transistor performance by increasing the threshold voltage V_T. Increasing the V_T of a transistor lowers the I_ON of the transistor which lowers the speed of the transistor. This is probably the reason why their processor could no longer overclock as high as before. Also running the CPU 24/7 aggravates the situation as more switching (electron activity) occurs which will increase the rate at which those trapped charges accumulate.

I think there is a common misconception that overvolting a CPU while keeping the temperature below a certain threshold won't damage the CPU. This is true to a certain extent because lowering the temperature helps meet the metal current density (so metal interconnects don't burn out due to the increased currents). But applying a high voltage to the CPU beyond the spec will certainly cause "hot electrons", which is something the CPU makers consider when setting the maximum voltage spec on their chips.

The bottom line is that overvolting will cause a degradation in the performance of the CPU over time. How long before you begin to notice the degradation is hard to tell.
 
Originally posted by: Markfw900
Its a brand new (2-3 months old, whole system was new) 700 watt Fortron. I can't believe its overloaded and failing. I don't have anything better to swap, maybe a Antec 450.

aigomorla: Its a P5K ASUS (non-deluxe). With those two Antec 900 fans blowing right over the motherboard, one pulling, and the 200mm pulling the other direction, I have plenty of airflow. I may need some ramsinks over the MOSFETS.
Click to expand...

mark. under normal conditions most game boards arent designed to be ran at full load 24/7 overclocked.

My P5K-E failed this way. You seriously need some sort of cooling on those mosfet for an overclocked quad crunching.

That evga thread only shows what can happen because of poor contact with cooling these mosfets down.

Try and see if you have good contact on those mosfets. And definitely put a sink on them. My personal favorate are these:

http://www.jab-tech.com/Enzote...nk-BMR-C1-pr-3724.html
 
Whats funny, is that beside my DQ6/G0 system, my best OC at 3200 is an ols S3 motherboard with a B3 stepping chip. Hasn't blinked in 3 months !! Set it at 8x400 and let it go ! The one having the problem now is running 435x7, but it has been over a day now, and stable.
 
ouch i hate to hear about something like that happening. are you able to maybe take that chip out and try running it on on of your other boards? this should verify if the problem is the board.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top