look like your Overclock is hitting a rough road huh? BSOD.... why dont you try backing down your cpu to 3.5 or 3.0 or even stock and see. I doubt its your powersupply. IT might just be your cpu not happy with running 4.1 or 4.0 ( its not a stable overclock afterall) It couldnt even pass 2 hours of prime. BTW, you really should find a speed where you can pass at least 4+ hours of prime and work your way up. instead of running at 4.1ghz and work your way DOWN to a stable speed.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Q6600 @ 4.1 GHZ
- Thread starter annaconda
- Start date
Alright i tried All the way down upto 3.6GHZ and i am not passing Street Test more than 2 hours.
lol, definately PSU, and it hapens when i tried to do multi-tasking while running Prime95.
Unfortunatly, that doesn't mean anything, it could be your CPU is not 100% stable even at 3.6ghz.
These days, just because you can boot a chip and run small tests at 4+ghz, doesnt mean that the actual stable 24/7 settings are anything close to that.
I've seen chips that can run some benchmarks at 4.2ghz, but yet won't pass OCCT or Prime 95 tests until they're lowered down to 3.4ghz or so.
There are a lot of factors too... it could be the memory is peaked out, it could be the northbridge is not able to keep up, hell, who knows for sure.
The power supply is probably being stressed pretty badly, however, I would be shocked it changing it out for a more powerful unit would yield you more stable results.
That is the way overclocking works, it's so hit or miss, there are so many variables.
I would suggest setting everything as stock as possible, (ie ram timings) and working your way up slowly. First start at 3ghz or so and Prime overnight, when you go to sleep... if it's still going the next day when you wake up, fantastic.
The next night, 3.2ghz, so on and so on.
Or run OCCT for 30 minutes at a time, that will stress things to their limits.
I'll bet you could work 24/7 stability at 3.6 to 3.8ghz without much trouble, anything past that I seriously doubt it.
lol, i failed PRIME95 @ Stock after 1 minute
Looks like i getting to the bottom of this. I think this is because of the NVIDIA's latest chipset driver. I Network Teaming acting weired, i am going to disable NetworkTeaming and try @ stock again.
What do you guys say now ?
Looks like i getting to the bottom of this. I think this is because of the NVIDIA's latest chipset driver. I Network Teaming acting weired, i am going to disable NetworkTeaming and try @ stock again.
What do you guys say now ?
lol, i failed PRIME95 @ Stock after 1 minute
Looks like i getting to the bottom of this. I think this is because of the NVIDIA's latest chipset driver. I Network Teaming acting weired, i am going to disable NetworkTeaming and try @ stock again.
What do you guys say now ?
I say it's time to flash your BIOS, format your hard drive and start all over.
But I'm dramatic that way.
Honestly though, it could be timings, pci clocks, who knows, why too many seperate issues at play here, this is the problem with stability testing, once you enter down that path, it's hard to know when to stop.
There is a chance that you fucked up your cpu by running it at 4 ghz even thou its not stable. How long you had your cpu at 4ghz+?? a week? few days?? you are running 1.5v doesnt help either. Hopfully its somehting else....rams, northbridge......good luck.
LOL, don't scare the poor dude, little chance of that, these new Intels are quite durable, if he'd been pumping 1.75v through there for weeks, I'd be worried, but at his volts, with his temps... even adjusting for the 15c, there is little chance he damaged anything.
More then likely he's tweaked a setting with his RAM that is keeping him from stability, or he might be running on an old copy of Windows, so many things could keep him from passing Prime95.
There are many reports of people on a stock system being unable to run stress tests, so the variables are many.
As I said, with tweaks I'll bet he could find 3.6ghz or even 3.8ghz stable for 24/7, anything past that is a pipe dream.
I'd be most worried about the motherboard. I've had interesting results on three boards using the Q6600 overclocked.
Gigabyte P965-DS3 with Q6600 @ 3.8GHz with about 1.45V real.
I ran prime for about 5 minutes before I noticed some bad fumes coming out of my computer. Immediately stopped prime and examined the board. No physical damage at all, but something was definitely overheating to the point that if I had kept running it nonstop - the board probably wouldn't work today anymore. End result was the board was fine.
Abit IP35-E with Q6600 @ 3.6GHz with about 1.35V real.
I ran prime, and while using the monitor, I noticed the PWM temp had risen to 100C under load and this caused the computer to shutdown (safety measure). Probably the same thing had happened on the DS3, except without the shutdown. Who knows how hot the PWM on the DS3 got. With a fan mounted on top of it, I was able to keep it under 100C primed, but not much. It maybe hit 95C... way to hot. Board had other non-related problems and was returned.
Gigabyte P35-DQ6 with Q6600 @ 3.6GHz with about 1.36V real.
This was prime stable for over 4 hours with the settings. Had a fan over the PWM area the entire time, and picked a night where my room temp dropped to about 68F. I was looking to minimize to my VCore. On my second run, the computer locked up after about 3 hours. I rebooted, and everything became unstable. All my old overclock settings were unstable. It was like I had lost 200-300MHz of stability at every voltage that was once hours stable. I thought I had damaged my Q6600, but it never passed 70C under load during these tests. I didn't know what was going on. The next morning, after the computer was off for a few hours, everything was back normal. Everything was rock solid again, as if the instability I was experiencing the night before never happened.
So, I've concluded it's not worth it to run a quad at extreme speeds for a long period of time. I do have 3510MHz proven stable at a low VCore in the O/C database, but given the circumstances I've recently experienced, 3200GHz seems like a "safe" point for me right now - especially given that I can run that at stock VCore. The OPs CPU is definitely fine.
Maybe in two years when the quad starts to feel slow, if it starts to feel slow, will I go for 3.6GHz again.
Gigabyte P965-DS3 with Q6600 @ 3.8GHz with about 1.45V real.
I ran prime for about 5 minutes before I noticed some bad fumes coming out of my computer. Immediately stopped prime and examined the board. No physical damage at all, but something was definitely overheating to the point that if I had kept running it nonstop - the board probably wouldn't work today anymore. End result was the board was fine.
Abit IP35-E with Q6600 @ 3.6GHz with about 1.35V real.
I ran prime, and while using the monitor, I noticed the PWM temp had risen to 100C under load and this caused the computer to shutdown (safety measure). Probably the same thing had happened on the DS3, except without the shutdown. Who knows how hot the PWM on the DS3 got. With a fan mounted on top of it, I was able to keep it under 100C primed, but not much. It maybe hit 95C... way to hot. Board had other non-related problems and was returned.
Gigabyte P35-DQ6 with Q6600 @ 3.6GHz with about 1.36V real.
This was prime stable for over 4 hours with the settings. Had a fan over the PWM area the entire time, and picked a night where my room temp dropped to about 68F. I was looking to minimize to my VCore. On my second run, the computer locked up after about 3 hours. I rebooted, and everything became unstable. All my old overclock settings were unstable. It was like I had lost 200-300MHz of stability at every voltage that was once hours stable. I thought I had damaged my Q6600, but it never passed 70C under load during these tests. I didn't know what was going on. The next morning, after the computer was off for a few hours, everything was back normal. Everything was rock solid again, as if the instability I was experiencing the night before never happened.
So, I've concluded it's not worth it to run a quad at extreme speeds for a long period of time. I do have 3510MHz proven stable at a low VCore in the O/C database, but given the circumstances I've recently experienced, 3200GHz seems like a "safe" point for me right now - especially given that I can run that at stock VCore. The OPs CPU is definitely fine.
Maybe in two years when the quad starts to feel slow, if it starts to feel slow, will I go for 3.6GHz again.
maybe it's time for a new thread.
we will call it:
"taking bets - how long will annaconda's speed-drunken 4.1GHz rampage continue before his cpu decides that it has had enough and it needs a break so then it will be time for tears?"
edit: i'm partly joking, of course. but truthfully, i think that bringing it back down to 3.8 or 3.9 and working from there would be smart. as it stands, you either need to kick down the speed or increase the voltage.
my advice: don't let there be any doubt as to the stability of a machine which you obviously seem to be very meticulous about. if anyone hasn't seen the setup that annaconda is working with then check it out, and tell me that you would come so far with a machine as he has to let it be semi-stable at this incredible speed.
make it stable and get it in the database!
we will call it:
"taking bets - how long will annaconda's speed-drunken 4.1GHz rampage continue before his cpu decides that it has had enough and it needs a break so then it will be time for tears?"
edit: i'm partly joking, of course. but truthfully, i think that bringing it back down to 3.8 or 3.9 and working from there would be smart. as it stands, you either need to kick down the speed or increase the voltage.
my advice: don't let there be any doubt as to the stability of a machine which you obviously seem to be very meticulous about. if anyone hasn't seen the setup that annaconda is working with then check it out, and tell me that you would come so far with a machine as he has to let it be semi-stable at this incredible speed.
make it stable and get it in the database!
just curious, what honestly is bottlenecked by the CPU at 3.0ghz that warrants an additional 1.1ghz? other than benchmarks maybe. (insert minesweeper joke here)
i didn't mean to sound like i was insisting the chip should be stable at 4GHz+.
i just meant that he should run the chip at whatever can be deemed highest stable via generally accepted stress testing methods, get the chip in the database, and leave it there.
zmjone2992
[H]ard|Gawd
- Joined
- Jul 3, 2007
- Messages
- 1,033
I believe dell ships their qx6850 chips at 1.6v. But those are their bin+3 sorted chips. So I suppose the 90% might apply there. I think they discontinued their highest overclock for just that reason actually. 3.73 if I recall correctly.
lol, i failed PRIME95 @ Stock after 1 minute
Looks like i getting to the bottom of this. I think this is because of the NVIDIA's latest chipset driver. I Network Teaming acting weired, i am going to disable NetworkTeaming and try @ stock again.
What do you guys say now ?
Dude, I seriously hope you didn't F*ck your chip somehow. That's a lot of dough down the toilet after the lapping. I doubt it, but it would still suck. If you can't pass at stock settings then something is definitely wrong. Have you memtest'ed your RAM? Start eliminating possibilities.
PS: You'll get over the speed difference. I actually am thinking about running my CPU full time at [email protected] since I honestly don't notice a huge difference between that and 3.6. I actually get some peace of mind from undervolting and running at 3.2ghz under 50C load after 8+hours. Gaming and most other tasks never bring this CPU above 40C even with an 85C ambient room temp.
That's the point i am trying to make. His cpu CANT make the database. to make the database, you ONLY NEED 4 hours of prime. This thread should be closed. I bet his chip cant even get 3.8 or 3.9 stable. 4.1 is NOTHING to brag about when you cant even get 2hours of prime stable let alone 4. you feed any G0 Q6600 with 1.6v + watercooling, you will get it to post at 4.0+ and run some benchmarks.
Wow, why you are so upset ? lol. Speak for your self when you say stop braging i am not braging about any thing. Who cares about the database ? you might do i don;t. You are saying i am bragging you are the one who want to show off being in the database.
I am only sharing my experience with others who care to help each other and learn. People like you create bad reputation for the forums. Oh sorry my last comment was wrong, actually you don;t represent the forum, and i don;t care.
NOW BACK TO DISCUSSION:
No chip is alright and it is perfectly running @ stock right now. I know it is PSU like i said so many times before, i check it with the Kill-A-Watt and it is drawing 430 watts, which i think could fluctuate during different torture test or unload.
Now i am have to wait for to buy a new PSU at least i think 850 watts to see how it goes. I did not tried the memory test, that would be good to. The reason i did not post any tests or results is because i have been busy studying for my CCNA, and looking for new JOB.
But thanx again for every one who helped and appreciated.
Well, as much as that forum member might annoy you, he's pretty much right. I had an IP35 Pro and it was only stable at 3.2ghz, but I could run superpi 1mb at 4ghz before it locked up.
The database is far from a bragging platform. It requires proof of mild stability and shows what hardware was used in achieving each overclock. I think it is an invaluable resource. Posting overclocks on here only shows that you can run superpi or a few other programs without locking up. People on here aren't going to really care about 4ghz without stability, because an unstable overclock is essentially unusable.
I will be very surprised if you get 4ghz stable even with a new PSU, I hope you do, but I am not holding my breath. Your 430W power draw is getting me worried. My Hiper Type R is rated at 580W, so I am at the limit if I am getting 80% efficiency.
The database is far from a bragging platform. It requires proof of mild stability and shows what hardware was used in achieving each overclock. I think it is an invaluable resource. Posting overclocks on here only shows that you can run superpi or a few other programs without locking up. People on here aren't going to really care about 4ghz without stability, because an unstable overclock is essentially unusable.
I will be very surprised if you get 4ghz stable even with a new PSU, I hope you do, but I am not holding my breath. Your 430W power draw is getting me worried. My Hiper Type R is rated at 580W, so I am at the limit if I am getting 80% efficiency.
Here's some rough estimates of what my chip does with power draw (and you see how significantly the power draw goes up once you leave stock voltage).
1.2375V = my stock stock voltage
2400Mhz - 1.2375V =~ 95W
3200MHz - 1.2375V =~ 127W
3600MHz - 1.3600V =~ 172W
4100MHz - 1.5125V =~ 242W (If I were to use your settings here)
I don't know about you, but there is no way in hell I would run that speed 24/7 even if under water. I'd maybe go for a 4hr stable setting, but I would bolt down all the heatsinks on the motherboard, have active cooling on all of them, along with the RAM. I would remove any PCI cards I don't need, use only one hard drive, and use a crappy video card like a 7300LE going for $10ARGC right now.
1.2375V = my stock stock voltage
2400Mhz - 1.2375V =~ 95W
3200MHz - 1.2375V =~ 127W
3600MHz - 1.3600V =~ 172W
4100MHz - 1.5125V =~ 242W (If I were to use your settings here)
I don't know about you, but there is no way in hell I would run that speed 24/7 even if under water. I'd maybe go for a 4hr stable setting, but I would bolt down all the heatsinks on the motherboard, have active cooling on all of them, along with the RAM. I would remove any PCI cards I don't need, use only one hard drive, and use a crappy video card like a 7300LE going for $10ARGC right now.
jaegertech
Gawd
- Joined
- May 10, 2005
- Messages
- 603
Uhm, do you think 80% efficiency means a PSU runs at 80% of rated capacity? The efficiency rating is for how efficiently the PSU converts AC to DC. In other words, a 580 watt psu running at 100% load and 80% efficiency is drawing 725 watts of AC from the wall. The 20% is presumably lost as heat. So highly efficient power supplies mean lower power bills and lower heat loads.
That Antec Quattro 850 keeps showing up at $130-150. I haven't seen anything else that approaches that level of value.
Thanks Dethred for saying i am right. He thinks i am annoying because he knows i am right. Showing screen shots at 4.1 without backing them up is childish. Starting a thread here AND IN XTREME SYSTEM with tons of pics and saying "perfectly stable @4.1" is begging for it. If the thread title is 4.1 semi stable in windows, no one would have said anything negative to him at all. I would probably said " good chip "
quote
"Trust me guys it is freakin fast @ 4.1. There is no way i can go back to stock or 3ghz."
If this is not bragging then i dont know what is. As for him saying that i brag about in the database, sorry kid...i am not in the database. i could care less if i am in the database. The point i am trying to make is its EASY to get 4+hours of prime and into the database. The fact that his cpu cant even make the 4hours database tell you how UNSTABLE IT IS.
quote
"Trust me guys it is freakin fast @ 4.1. There is no way i can go back to stock or 3ghz."
If this is not bragging then i dont know what is. As for him saying that i brag about in the database, sorry kid...i am not in the database. i could care less if i am in the database. The point i am trying to make is its EASY to get 4+hours of prime and into the database. The fact that his cpu cant even make the 4hours database tell you how UNSTABLE IT IS.
When i say perfectly stable i mean, for my daily usage. Let me give you example of my daily usage.
Heres what i tried, and it did not crashed.
VMWARE = 3 Operating Systems running all togather.
8GB Of Winrar File extracted.
WebSurfing.
Watching DVD.
I know i cannot pass 4 hours of Prime95, but then again why would i care? Why would i do it? Looks like you are not getting it are you ?
I mean who use PRIME95 every day for 4 hours ?
I listened to UncleWeb, and Turtletrax and did the MOD on Motherboard and reduced the VCORE @ 1.512 @ 4050, and it is perfectly running since last week.
If you are saying your can also do that then good for you. Go ahead post your screen shots and prove it.
The reason i have this post is to get help and feedback on the experience i am having. People really helped they pointed out some good points, like keep your voltage under 1.55v, and do the MOD to the motherboard.
Now just tell me why would i need to run PRIME95 ? why ? why ?
Heres what i tried, and it did not crashed.
VMWARE = 3 Operating Systems running all togather.
8GB Of Winrar File extracted.
WebSurfing.
Watching DVD.
I know i cannot pass 4 hours of Prime95, but then again why would i care? Why would i do it? Looks like you are not getting it are you ?
I mean who use PRIME95 every day for 4 hours ?
I listened to UncleWeb, and Turtletrax and did the MOD on Motherboard and reduced the VCORE @ 1.512 @ 4050, and it is perfectly running since last week.
If you are saying your can also do that then good for you. Go ahead post your screen shots and prove it.
The reason i have this post is to get help and feedback on the experience i am having. People really helped they pointed out some good points, like keep your voltage under 1.55v, and do the MOD to the motherboard.
Now just tell me why would i need to run PRIME95 ? why ? why ?
Damn, you are right....nobody runs prime for 4+ hours or overnight EVERYDAY. But almost everyone run it ONCE to at least know that their computer is 'somewhat stable'. If you could pass prime, you would have post pic of it instead of all this bullshit about who needs to run prime blah blah blah.. Ask everyone...how many people have run prime 4+ hours and i am sure at one time or another everyone DID. and yet nobody uses prime everyday...do the math. I am sick answering you. keep bragging about it, nobody give a shit about any overclock if its NOT STABLE. If people want to see unstable overclock, they can go to Xtreme system forum and see 5.0ghz screenshots.
P.s i dont need to prove shit to you...i am not the one to start a bragging thread
P.s i dont need to prove shit to you...i am not the one to start a bragging thread
I've been lurking in this thread for some time and have seen valid arguments from all sides. Bottom line, if the above quote is true, that's all that matters at the end of the day. When I get a new system I never run things like Prime or what not to test for stability. I run the programs I use all the time, until I see an anomaly crop up. That's when I know to either 'drop it' or 'cool it.' If you're enjoying fantastic performance with your current settings with a host of commonly used applications without a single incident, then great. You've successfully accomplished your mission's objective. Employ what works for you.
TheDarkTao
Gawd
- Joined
- Feb 5, 2005
- Messages
- 800
I had a rig once (an old p3) it would never pass prime past an hour... I had the thing for over a year and no problems! Nothing!
I had an AMD opteron 146... great chip great OC. It also would NEVER pass prime. Ran em both at speeds prime NON-stable... Never had a problem. So that rule is not law that a system MUST be prime stable. I have an e6300 that will pass prime now for 10 hours... I get random lockups n what not from time to time (also vista/nvidia problems). But it is prime stable! (10+ hours)
Once in fact I primed it for 7 hours, then stopped and about 30 seconds later on opening IE the rig crashed
Just goes to show, prime is not law
Enjoy the rig, and just fwiw in over 20 rigs I have yet to have one issues since the p3 days with a CPU being destroyed from overclocking!
I had an AMD opteron 146... great chip great OC. It also would NEVER pass prime. Ran em both at speeds prime NON-stable... Never had a problem. So that rule is not law that a system MUST be prime stable. I have an e6300 that will pass prime now for 10 hours... I get random lockups n what not from time to time (also vista/nvidia problems). But it is prime stable! (10+ hours)
Once in fact I primed it for 7 hours, then stopped and about 30 seconds later on opening IE the rig crashed
Just goes to show, prime is not law
Enjoy the rig, and just fwiw in over 20 rigs I have yet to have one issues since the p3 days with a CPU being destroyed from overclocking!
Guys guys guys. This isn't a pissing match here. This is a simple argument over what is stable and what is not. I can see that annaconda is seriously looking at improving his machine and is asking questions, but at the same time he's probably proud that he has a quadcore up to that speed and it works for his purposes.
To Annaconda his CPU is stable. Unfortunately, most overclockers, including myself, consider stress testing programs to be a great way of making sure their overclocked CPU is going to handle anything they throw at it and not crash the system. So Annaconda might need to reiterate what his definition of stable is and specify why his overclock works for him, otherwise people such as myself are going to dismiss it as an overclock that many CPU's (I'm saying this in a relative sense) can achieve with similar "stability".
That being said, there is no definitive stability measure. Some people stress test their machines for 24 hours and consider that stable, some do it for 4 hours and consider it stable. I personally do 8-10 hours because that is the longest amount of time I have ever or will ever do a CPU intensive task at one time. I have honestly found that if a stress test does not yeild errors within 30 minutes then the chances of failure are pretty slim. For all but one of my previous CPU's that I have overclocked, if the Stress test made it past 30 minutes it would pass for over 10 hours as well.
Not to be shallow an use another car analogy, but if someone turbocharged their car and doubled the power of the motor and claimed it was a daily driver, some might be skeptical. While the car's owner might drive it a mile a day on a secluded road and still call it a daily driver, others might call BS and point out that it can't do many things a grocery getter honda or likewise might be able to do.
To Annaconda his CPU is stable. Unfortunately, most overclockers, including myself, consider stress testing programs to be a great way of making sure their overclocked CPU is going to handle anything they throw at it and not crash the system. So Annaconda might need to reiterate what his definition of stable is and specify why his overclock works for him, otherwise people such as myself are going to dismiss it as an overclock that many CPU's (I'm saying this in a relative sense) can achieve with similar "stability".
That being said, there is no definitive stability measure. Some people stress test their machines for 24 hours and consider that stable, some do it for 4 hours and consider it stable. I personally do 8-10 hours because that is the longest amount of time I have ever or will ever do a CPU intensive task at one time. I have honestly found that if a stress test does not yeild errors within 30 minutes then the chances of failure are pretty slim. For all but one of my previous CPU's that I have overclocked, if the Stress test made it past 30 minutes it would pass for over 10 hours as well.
Not to be shallow an use another car analogy, but if someone turbocharged their car and doubled the power of the motor and claimed it was a daily driver, some might be skeptical. While the car's owner might drive it a mile a day on a secluded road and still call it a daily driver, others might call BS and point out that it can't do many things a grocery getter honda or likewise might be able to do.
I had a rig once (an old p3) it would never pass prime past an hour... I had the thing for over a year and no problems! Nothing!
I had an AMD opteron 146... great chip great OC. It also would NEVER pass prime. Ran em both at speeds prime NON-stable... Never had a problem. So that rule is not law that a system MUST be prime stable. I have an e6300 that will pass prime now for 10 hours... I get random lockups n what not from time to time (also vista/nvidia problems). But it is prime stable! (10+ hours)
Once in fact I primed it for 7 hours, then stopped and about 30 seconds later on opening IE the rig crashed
Just goes to show, prime is not law
Enjoy the rig, and just fwiw in over 20 rigs I have yet to have one issues since the p3 days with a CPU being destroyed from overclocking!
You aren't assuming that because there is an instability that its, by default, a hardware issue, are you? Software problems with some configurations are sometimes unavoidable, but adding in hardware malfunctions or instability will compound your headaches.
I don't think it really matters to most people. I think the few people that dismiss someone else's acomplishments because they don't measure up to their standards are not worth listening to anyways.
Stable to most people means they can run any program at any length of time and it won't produce a hardware fault. While running a select few programs with relative stability will suffice for now, there is a time when new programs or tasks may need to be run and the CPU can't handle it. I believe that the other poster in this thread is essentially saying, that if the CPU can't be known to run as well and stable as at stock speeds, then its not 100% stable.
Again, that is up to the owner to decide. Its all relative anyways.
Again, that is up to the owner to decide. Its all relative anyways.
winston856
[H]ard|Gawd
- Joined
- Dec 28, 2004
- Messages
- 1,822
Still on that crappy PSU are you?
geez some of you people are straight up assh0les
not everyone gives a flying fvck about your little database.... from what ive read so far, annaconda sure doesnt.
if this was the OC forum, maybe you'd have some foundation for your bitching and moaning about what kind of threads can/cannot be made... but this is the Intel GENERAL forum.. and You people are NOT moderators. Go back to staring at your name in the OC database and let everyone else enjoy their rigs as they please
not everyone gives a flying fvck about your little database.... from what ive read so far, annaconda sure doesnt.
if this was the OC forum, maybe you'd have some foundation for your bitching and moaning about what kind of threads can/cannot be made... but this is the Intel GENERAL forum.. and You people are NOT moderators. Go back to staring at your name in the OC database and let everyone else enjoy their rigs as they please