failed OC

[Rigit]

So I've got my rig stable at 4Ghz. Ran prime95 for couple hours no issues. When I went for 4.1 after a minute or so cpu 2 threw an error and prime shut it down. The test then continued for about 10 minutes and I got a BSOD. Core 2 likely caused the failure. In bios there's a setting for each core. I don't think I can change voltage on individual cores but it's a -2% setting that can be changed. I'm not sure but this would seem to indicate that I can reduce the stress on 1 core while leaving the others to run? Would reducing the load help? Do the other cores have to pick up the slack? Voltage for the attempt was 1.5. Not acceptable for full time use but I would like the trophy. CPUNB was 2600mhz and the HT was 2200. CPUNB voltage was 1.3 but for 2600 it shouldn't need to be that high. Ram was at factory spec with the voltage bumped 1 notch which does fine at 4.1 100mhz doesn't mean squate for everyday use. Wouldn't be at all noticeable but just to get there....For the record I don't like prime95 i don't think it's a fair test. It is extremely unlikely that I'll find a program that uses all four cores 100% running 4 threads each. It's just not a real world test. Even with all of the video editing/converting I do I don't think that happens. Most programs just aren't written to do it.

 

[RGone]

You don't think what we normally use as a stress testing application is any good so maybe what we think is no good also? Humh?

ACC with the core plus and minuses was released by AMD for the Phenom 1 cpus on 7xx chipset. No idea what the results of using it on the Phenom 2 will be. Try it and see is the only way to know what happens with your computer.

Not knowing how much stress was being applied since you detest P95, it would be my suspicion that your cpu just does not have enough Vcore to touch the magic 4.1Ghz. If the core 2 failure happened at much less than a full load then the Vcore must really not be high enough for the 4.1Ghz speeds. That maybe the best guess for the current situation.
RGone...

 

[DustyWolfe]

Corsair H50 Liquid CPU cooling with dual 120mm fans in push/pull configuration on the rad

what kind of temps are you seeing? that probably isn't gonna be enough cooler for the goals you seem to have

 

[Johan45]

Posted by Rigit
For the record I don't like prime95 i don't think it's a fair test. It is extremely unlikely that I'll find a program that uses all four cores 100% running 4 threads each. It's just not a real world test. Even with all of the video editing/converting I do I don't think that happens. Most programs just aren't written to do it.

That really depends on what you're doing.There are guys that run F@H that will root out instability where P95 doeesn't. I use some Bluray software that uses 8 cores plus both GPU's it's like a micro furnace when I'm crunching and 1 error I have a coaster.
I'm with Rgone sometimes it takes another .05v to get that 100MHz. that's " The Wall" you need to decide if it's worth it !

 

[Rigit]

You don't think what use as a normal stress testing application is any good so maybe what we think is no good also? Humh?

ACC with the core plus and minuses was released by AMD for the Phenom 1 cpus on 7xx chipset. No idea what the results of using it on the Phenom 2 will be. Try it and see is the only way to know what happens with your computer.

Not knowing how much stress was being applied since you detest P95, it would be my suspicion that your cpu just does not have enough Vcore to touch the magic 4.1Ghz. If the core 2 failure happened at much less than a full load then the Vcore must really not be high enough for the 4.1Ghz speeds. That maybe the best guess for the current situation.
RGone...

Don't know why you take such offense to my not liking p95. Whatever. If you'd read my post then you'd know I DID use p95 for the stress test because it does seem to be the gold standard. Therefor it was at full p95 load. I never said it or you were no good. I only said it doesn't simulate real world use. I'm using the p95 test results to trouble shoot and perhaps make adjustments for another try. There's nothing wrong with the data it gave me and if I felt the guys here were no good then I'd have never posted in the first place. I was unaware that the rules say I can't disagree with you or anyone else for that matter.

 

[Rigit]

what kind of temps are you seeing? that probably isn't gonna be enough cooler for the goals you seem to have

Temp never got higher than 49C. More voltage is likely needed for 4.1Ghz. So I'll likely not be trying again since that high of voltage is unsafe for everyday use and 100Mhz just isn't worth it to me.

 

[DustyWolfe]

I don't see any offense taken in his reply...looks like good solid info if you asked me. he is just trying to work around your "dislike" for prime I don't see the problem here

depends on your definition of safe..... I run 1.45 @ 4.1 ghz and 1.54 at 4.3ghz when ambients allow I suppose this isn't a one size fits all concept though. the way I see it as long as temps are fine why not

 

[Rigit]

That really depends on what you're doing.There are guys that run F@H that will root out instability where P95 doeesn't. I use some Bluray software that uses 8 cores plus both GPU's it's like a micro furnace when I'm crunching and 1 error I have a coaster.
I'm with Rgone sometimes it takes another .05v to get that 100MHz. that's " The Wall" you need to decide if it's worth it !

So not liking/using p95 doesn't make me a traitorous bas...? I'm not really trying for the "trophy" per say. I chose my words poorly. Stable to me means I can do everything I use my box for at whatever speed and have no problems. Now if I get into a competition here then yea I'd have to use p95 to prove stability. But I'm not competing with anybody and I've heard other guys express the same opinion as I and not get flamed. The extra 100Mhz isn't worth it. 4.0 is stable and I used p95 to prove it. I've been a member here a long time and this is the first time this has happened to me here. This is the greatest tech/oc'ing site out there. Folks are friendly, knowledgeable and eager to help. Sure hope this thread is an exception. I also hope it gets closed.

 

[DustyWolfe]

well..... that escalated quickly.

 

[RGone]

Actually I take no offense if one does not believe much in Prime 95 Blend mode. About the only reason to run Prime 95 is to try and put a great load on the system and determine if there appears to be any instability. I have run P95 Blend at 5.0Ghz on my FX-8350 and a passing thru without issue, leads me to believe I will be pretty stable at my normal speed of 4.0Ghz for general use and 4.8Ghz for video editting. I have used Lin-X that many declare is better than P95 Blend and can pass Lin-X at a lower Vcore than I must have for P95 Blend mode with every other setting the same but for the cpu voltage. So I can see why many would prefer Lin-X.

Over the last couple of weeks I have been working with slimming down Win 7 installation media. After all the changes I wish to see in the install media are hand setup, then the program begins the work of doing as I requested. During the majority of that phase it is difficult to open CPUz or HWMonitor or the Internet, so I suspect the cpu cores are pretty well loaded. Heck no I do not wish a failure during the slimming of my Win 7 media. It is enough trouble to get right anyway. But knowing that I am P95 stable gives me the confidence to know I need not expect a real problem under the apparent load being applied to my system while I am doing the work I wish to accomplish.

I guess it depends on where we are standing when we view a situation.
RGone...

 

[Rigit]

"You don't think what we normally use as a stress testing application is any good so maybe what we think is no good also? Humh? "
I guess I took that wrong. As for testing, I'll still use p95 since I don't know of any better test. There's no "real world" test I'm aware of. As for my cooler. I'm very anal about keeping it clean. I don't just blow it out with a can of air. Well I do but about once a week I pull the pull fan on the inside off and use a soft horsehair paint brush to get better cleaning than air alone. About every 3 months or so I just take the whole thing off and wash the rad, blow it dry with compressed air and let it sit for several hours to make sure it's dry before reinstalling it. I typically don't keep the side on my case either. As for temps, about 29C is typical @ idol but with summer approaching ambient air will be warmer and that will change. As for Vcore, the Vcore for 4.1 test was 1.5. Max temp was 49C. There seems to be a lot of opinions. I've read 1.5 is bad for continuous use. In this thread not so much. If I try again I'll bump the vcore. Funny thing though. I couldn't get past 3.8 with p95 until I installed my SSD.

 

[DustyWolfe]

honestly I couldn't tell a difference be it in fps or everyday computing between 4.0 and 4.3 ghz. I feel like the only reason I had it up that high was to say I did. I think you reach a point of diminishing returns after some point and you need to make a decision if that extra hundred or so MHz is actually gonna be worth it or not. for me I decided in the interest of not cleaning my rad twice a month id just crank her down a bit

 

[Rigit]

honestly I couldn't tell a difference be it in fps or everyday computing between 4.0 and 4.3 ghz. I feel like the only reason I had it up that high was to say I did. I think you reach a point of diminishing returns after some point and you need to make a decision if that extra hundred or so MHz is actually gonna be worth it or not. for me I decided in the interest of not cleaning my rad twice a month id just crank her down a bit

Yea, I didn't expect "blown away" results from a meager 100Mhz. It was getting there and I do a bit of video editing. All DVD's are digitized for AppleTV (a gift from my son. I'd have preferred a non-Apple device since Apple doesn't play nice with PC) Doing one DVD is easy enough. But at times I have 4 encoders going at once for various reasons. Pushes all 4 cores to 100% and doesn't let up until the jobs start completing. So I'll take all the power I can get. I did a test test with p95 @ 4.1Ghz but used the first test that doesn't test much memory. Voltage was bumped to 1.525 ran for an hour. No problems. I suspect if I run a blended test, it will crash. I've got Crucial 7-7-7-20 1066 memory. I think if I upgraded the memory I could go farther. But that's another forum. I used to be a mid-range OCer. Getting to a point where I said well enough. Learning to go hard core has proven more difficult then anticipated. 1 seemingly insignificant setting can make or break. Maddening.

 

[Mandrake4565]

rigit I had a look at your motherboard and I see one reason why you may be having difficulty getting higher overclocks especially with AIO water. Your VRM section has no heatsinks on it, this is ok for the average user but when overclocking you are putting more stress on this section. Also because of the AIO you probably have less air flow over this section then you would have if you were using an Air cooler. Possible fixes for this could be adding heatsinks to this section and or putting some airflow on this section. When I was pushing my 955be over 4.2 on water with a motherboard that had heatsinks on this section I still had to add a small fan blowing down on the heatsink in order to cool that area. See the area I highlighted.

 

[Johan45]

Hey Rigit wasn't trying to step on your toes. I just wanted to point out that P95 isn't the toughest program out there and some users appreciate the stability. I understand that cranking things up for gaming isn't quite the same as some of the things that others use their systems for.

 

[Rigit]

rigit I had a look at your motherboard and I see one reason why you may be having difficulty getting higher overclocks especially with AIO water. Your VRM section has no heatsinks on it, this is ok for the average user but when overclocking you are putting more stress on this section. Also because of the AIO you probably have less air flow over this section then you would have if you were using an Air cooler. Possible fixes for this could be adding heatsinks to this section and or putting some airflow on this section. When I was pushing my 955be over 4.2 on water with a motherboard that had heatsinks on this section I still had to add a small fan blowing down on the heatsink in order to cool that area. See the area I highlighted.

Thanks Mandrake that's good info. Next major upgrade I ask around here for a suitable board for my needs. However, I have 2 fans on the rad push/pull config drawing air in and over the area in question. I felt those VRM and they're warm to the touch by by no means hot. But then at the moment the side is off and a fan is blowing in which normally isn't the case. I do tend to leave the side off. Leaving the small personal fan blowing on it isn't any trouble if I need it. I could rig up a louver on the inside fan that directs some air down on VRM but it would be warm since it's coming through the rad. Right now I really think memory is hurting me. It can pass prime when not using blended. So leave memory out and pass. Put memory into the equation and it fails. I'm going to consult Dulk's guide and try again. Dulk's guide is the coolest thing since the heat sink as far as Phenom II's go.

 

[RGone]

Hey Rigit wasn't trying to step on your toes. I just wanted to point out that P95 isn't the toughest program out there and some users appreciate the stability. I understand that cranking things up for gaming isn't quite the same as some of the things that others use their systems for.

Man you said a mouthful there. I was cruising just a few days ago about testing system stability and I saw one person say that he did not consider his system stable unless it could run P95 and Furmark at the same time without issue. I slammed myself in the forehead because I could visualize the heat inside that case.

I used to be inching up on a max stable setup and during the stressing would open Internet Exploder. Often instant fubar. Knew I was not as stable as I thought. Hehehe. I worked from home for over 6.5 years and had emails and addresses and other necessary information and never overclocked that rig. Much more knowledgable IT persons just said it was not a good idea that over the course of time I would have enough issues keeping my data safe.

They were correct. I had two sets of ram fail and one hard drive fail and without having also done what they advised which was a religious-like back-up scheme, I would have been in serious deep doo doo. It all boils down to what one says he can live with I guess.
RGone...

 

[Mandrake4565]

How many sticks of ram are you running? 4x1g will put a bigger strain on the IMC then 2x2g. Have your tried running memtest86+ at stock clocks to see if you are getting any errors from the sticks?

 

[storm-chaser]

For the record I don't like prime95 i don't think it's a fair test. It is extremely unlikely that I'll find a program that uses all four cores 100% running 4 threads each. It's just not a real world test. Even with all of the video editing/converting I do I don't think that happens. Most programs just aren't written to do it.

Well said, Rigit, I couldn't agree more. Its quite aggressive for personal use. Running prime95 for a two hour marathon has the potential to put huge restrictions on an overclock and in many cases this will limit what your system is capable of in the real world. These virtual limitations, if you will, imposed by the aggressive nature of the test leave what is called a grey area -- and that grey area can really be thought of as the elephant in the room. Or the Chat room, as in this case. Many, many times, the first answer to a new thread is: "Did you run prime for two hours to make sure your overclock is stable?" And let me be clear, that's a great question but the term 'stable' is qualified here with both "two hours" and "prime95" and that's where we start running into trouble; you see for some overclocking enthusiasts prime95 can be thought of as the quintessential instrument for measuring reliability. Without a pass, your overclock is useless, unfit and unreliable. And if you dare question it, you are not only questioning the program, but also questioning the person behind the name. You are calling into doubt the very essence of what that individual believes in and holds true when crafting the perfect overclock. What these folks forget, is variety, the spice of life, the simple point that not everyone who overclocks requires the same treatment. Not every user is running F@H 24/7. Not every user is a hard core gamer playing Crysis 3 for hours on end. Not everyone is using a PC in a resource intensive way and many of them will never, ever peg all CPU cores and subsystems for 2 hours straight. In reality, many users will never need the stability that prime95 demands. Case in point: I have seen a handful of overclocked systems fail in blend mode after 30 minutes of testing, yet maintain a record of 100% uptime for years, even during intensive use. How can this be? Because the reliability point measured in prime95 is much higher than what most users will ever be able to throw at it. Indicating there is a grey area between real world stability and that which is dictated by Prime95. And as a result, perhaps people are misguided into setting overclock limitations prematurely. This means they've lost out, for the sake of an over-aggressive reliability test. I want to push my system to the absolute limit of what it can handle in the real world, not in the world of virtual reality. I test stability by actually using the computer day by day, in the real world. In a very basic analogy, this entire issue of reliability can be compared to a sea trial for a new ship. You take it through countless tests of engine power and weather to determine if it is fit for its purpose and use on the open seas. You certainly don't throw a hurricane and a rogue wave at a 50 car ferry at the same time and expect it to stand tall. You tailor or harmonize the sea trials and reliability tests towards what that boat may be expected to encounter during its lifetime. Same thought process can be applied to reliability and how it is measured on a PC.

 

[Mandrake4565]

Storm I am not in 100% disagreement with you that prime95 2 hours is a little overkill for some users. With that said when we have people come here and ask us how to overclock and or help them try to stabilize there pc at high OC's. We don't know what they are using the computer for most of the time. Nor can we sit down at their desk and work on their pc, to help figure out what exactly is causing the instability, while trying to push on their cpu. As you well know overclocking is all about learning, a lot of people don't want to learn, they just want to come here and find numbers to plug into their system and viola they have an overclocked pc. You know as well as I do that it doesn't work that way. The ones who come here that want to learn and understand the inns and outs of overclocking will learn what their system needs to be stable for their uses. Some will find prime overkill some will find it a necessity.