Sign in with
Sign up | Sign in
Your question

Has anyone OCed a Core 2 Duo till it failed?

Last response: in CPUs
Share

Would you do it if you had the money to do so?

Total: 58 votes

  • Hells yeah, that''s the point
  • 56 %
  • Are you nuts? Stay within the limits of the chip.
  • 45 %
October 15, 2006 7:44:26 PM

The thread title is pretty self explanatory, I just wan to know if any of us has pushed the Core 2 to the point where it just won't work anymore.
I'd like to know if you have, if you would, where was the limit and what you think the limit for the Core 2 Duo is.
October 15, 2006 7:55:52 PM

i wouldn't overclock it to the max even if i had the memory because i dont have the money :(  to ensure that if i did fry the processor i could go out and buy another.

Though If i did have the money, perhaps yes, i would:) , I guess the max overclock would be somewhere around 5.5~6ghz, due to the short(er) pipeline....

------

I'm getting an intel e6600! yey! Tomorrow! (and zalman 7700alcu) :D 
a b à CPUs
October 15, 2006 7:56:02 PM

Haven't reached a limit cuz I haven't really pushed it to it's maximum. I've pushed it to 4.7GHz but not more. I just have bad memories of Northwood SDS (Sudden Death Syndrom) that was caused by electrical migration.
Related resources
October 15, 2006 8:00:27 PM

Highest OC on any C2D I have seen is 5525mhz on an X6800 by coolaler. If I had money, I would get an E6300 and see what she could take :) 
October 15, 2006 8:04:09 PM

can't you pin/volt mod it to get a higher voltage? :?
a b à CPUs
October 15, 2006 8:06:42 PM

Quote:
Happens for Core2 at beyond 1.55V which starts going at an unacceptable rate once 1.675V is reached.
Damn Intel built its protection from n00bz too well.


Not sure it's an intended Intel protection though. At least with Intel CPU's you know their process technology is superior to SOI. I mean Intel processors overclock like crazy where as AMD processors are always sorta near their limits.

If you look at it, although Intel Shrunk the amount of pipelines in the C2D as compared to the Prescott it still overclocks nearly as well as a Netburst processor.
October 15, 2006 8:08:52 PM

Quote:
The thread title is pretty self explanatory, I just wan to know if any of us has pushed the Core 2 to the point where it just won't work anymore.
I'd like to know if you have, if you would, where was the limit and what you think the limit for the Core 2 Duo is.
I wouldn't want to see how high it will go before death, because there's no gaurantee that the replacement chip will overclock to near the same speed....which could lead to 2 dead chips. I guess if i was a millionaire, maybe it would be interesting. :?
October 15, 2006 8:20:42 PM

I want to do this with a E6300, but i just wanted too see if there was a level where i could bring it too and see how much i could incrementally up it before it just died.
October 15, 2006 8:54:57 PM

one question, are you rich by any chance? or do you just have rich parents OR do just spend loadsa money on computers and have a good job :D :D :D  ?

lol
October 15, 2006 8:57:16 PM

kinda, yes, yes, no... :lol: 
I love this stuff as much as the next dude. I just try not to be the typical rich pamby boy that forgets where his roots are and needs to get his ass knocked back to earth.
October 15, 2006 9:26:27 PM

YOU. ARE. GOD.
October 15, 2006 9:39:27 PM

Hmmm if I had a core2 cpu......................I am not that big into over clocking. When I do over clock....and I have, it is usually near the end of the usefullness of the cpu to get extra power out of it.

Ninja I assume that you mean you want to over clock a core2 cpu to its workable limit or over clock it until it fries?

I have always hoped that that Tom's would do an overclock CPU chart. One chart air cooled the other cooled by other means.

It seemes to me that intell cpu's do overclock better than AMD cpu's but they don't seem to give you the same boost per overclock as an overclocked AMD cpu. Core2 being the exception.

Try Google I am sure you can find plenty of over clocking info on the core2 chips. I took a look and there were numerous articles.
October 15, 2006 10:12:59 PM

Quote:
YOU. ARE. GOD.

No Wusy is. I am merely his humble prophet and priest. Bow to the temple of Wusy.... Ohmmmm.
October 15, 2006 10:20:29 PM

Quote:
YOU. ARE. GOD.

No Wusy is. I am merely his humble prophet and priest. Bow to the temple of Wusy.... Ohmmmm.

Don't you try and take my position you jerk! I will lay the smack down on your ass!!!! lol j/k

In any case, I will get a C2D and I will OC it until something fails. Then I will repeat, compensating for the broken hardware. I have an audit internship coming, so I will definitely be able to afford a C2D, I am thinking E6600, maybe 6700 + RD600 (ty wusy) + TEC cooling.

I don't know the meaning of limits other than it is my job to break them 8O
October 15, 2006 10:22:01 PM

HAIL THE MAGICAL WUSY! HAIL!!!
October 15, 2006 10:22:21 PM

Quote:
YOU. ARE. GOD.

No Wusy is. I am merely his humble prophet and priest. Bow to the temple of Wusy.... Ohmmmm.Undead priest?
October 15, 2006 10:47:32 PM

Peltier cooling is cheaper, I know, but why not go for a Mach II GT?
October 15, 2006 11:07:21 PM

Because it is cheaper, and more adaptable. I am in college so I have to prioritize. the incremental cost of peltier over already existing water is much smaller then a complete redesign off of phase change, and that is my main concern, cost.

Edit: nice location... got any spare liquid nitrogen I can have?
October 17, 2006 11:04:26 PM

Yeah, that's fine. I was hoping somebody would push a C2D on phase change and provide some numbers here.

Nope, no LN, but I do have an unlimited supply of R404a...

BTW: Which peltier device are you using?
October 17, 2006 11:32:50 PM

Send me your C2D processor and I will let you know :lol: 
October 17, 2006 11:45:38 PM

Yo Ninja, where is ur zeon3060...you know who you want to give your x6800 to. :wink: :wink:

If the x6800 is still better than your zeon3060....you know who you want to give your zeon3060 to. :wink: :wink:

Best TEC peltiers....
VIGOR GAMING Monsoon II (the one Ninja is going to donate to me :wink: )

Titan Amanda TEC
October 18, 2006 12:00:53 AM

I think the scarcity of stories about fried C2D's comes from two factors.

1) For a given cooling method, there's a certain voltage beyond which you can't seem to stabilize higher frequencies. As you keep increasing the voltage, sometimes the overclock worsens, which tells you at once not to keep increasing voltage. The chip continues to operate at peak frequency at this optimal voltage, and even if the peak wears down a few percent due to overheating, it's still not a dead chip.

2) C2D has only been out for a few months, and most purchases are even more recent. Overclockers are careful to monitor temperatures, as you need a certain combination of voltage and temperature for a chip to slowly deteriorate. There may already exist such conditions, but we'll need time for the reports to start coming in.
October 18, 2006 3:32:21 AM

I am going to use a switchtech TEC, MCW5002-775T™ LIQUID COOLED THERMOELECTRIC ASSEMBLY FOR INTEL® PENTIUM® 4 (LGA775) PROCESSORS probably. It should provide much better cooling than water, of course much better is a relative term.

Talk to me around Christmas time... it should be broken by then :tongue:

What is R404a.... sounds dangerous and something I want to be a part of.

-3lfk1ng

Those are kinda weak peltiers.... the really good ones are just a copper square with bare leads and require an entirely separate PSU to power it. I am looking to stick a nice 226w peltier on my C2D, and it may get cranked up beyond 240w, if I get a meanwell PSU with voltage adjustment.
October 18, 2006 5:29:43 AM

Quote:
I'll be only buying the parts. The actual construction/gasing will be done by an AC repair company.
I've briefly looked into equipment required for building a phasechange without help, minimum safety and it ain't cheap...

Known Target: 1/10th of commercial phasechange price with better performance


Isn't it worth just avoiding all the hassle and get an astek vapochill ls?
October 18, 2006 7:44:24 AM

Quote:
I'll be only buying the parts. The actual construction/gasing will be done by an AC repair company.
I've briefly looked into equipment required for building a phasechange without help, minimum safety and it ain't cheap...

Known Target: 1/10th of commercial phasechange price with better performance


You've hit the nail on the head! Buying a stock unit like the LS is freakishly expensive, I can get 50,000BTU (sounds crazy!!!) of conventional cooling for that kind of money.

The fact that you've found an AC company to do the build work for you is an ENORMOUS plus. It's not impossible to do it yourself, granted you are familiar with AC equipment, but I know the average PC-enthusiast-slash-college-goer is NO AC technician.

The only thing I DON'T like about phase change is that you will need to eventually refill the unit with new coolant, and as you said, R404a ain't cheap and your compressor must be designed for it. R134a on the other hand I can get for just a few bucks from any auto parts store in ready-to-use cans.

I would look forward to the day where the thermal envelope of CPUs goes back down to the point where phase change cooling won't be required, but I can keep dreaming!!! Currently I'm on phase change, and I'll be staying with it for a while to maximize my investment, but for my next major build I will most likely go over to TEC/water because it is MUCH easier to put together and maintain.

Good luck with your AC build, I hope it exceeds your expectations!!!

Superfly, the Swiftech is an excellent choice, and the kit version comes with the meanwell psu. They build solid products.
October 18, 2006 12:17:27 PM

Jack, quick question

how have you managed to accrue over 5000 posts having only joined earlier this year!

You must live on these forums
October 18, 2006 1:23:08 PM

Quote:
Jack, quick question

how have you managed to accrue over 5000 posts having only joined earlier this year!

You must live on these forums
He's very knowledgeable, and likes to enrich the forums with his input. :) 
October 18, 2006 2:08:01 PM

Quote:
Jack, quick question

how have you managed to accrue over 5000 posts having only joined earlier this year!

You must live on these forums
He's very knowledgeable, and likes to enrich the forums with his input. :) 

So thats a yes, he spends too much time here..... lol j/k

I wish I had more time to devote here, but my schedule is nowhere near stable.
October 18, 2006 2:21:36 PM

Does that mean you are only half as knowledgeable as Jack, as you only have half the number of posts for a similar time frame??

just kidding!!

i'm not knocking anyone here

It made me think though, as i have actually been visiting tomshardware since about 1997 but only signed up this year too. i could'nt possibly rack up even your number of posts let alone Jack's and wusy's as i work too many hours have have loads of commitments.

OP: interesting question. but i can't afford to try and find out, i have almost as big a battle trying to keep my rev c0 prescott cool
October 18, 2006 2:25:31 PM

Quote:
Does that mean you are only half as knowledgeable as Jack, as you only have half the number of posts for a similar time frame??

just kidding!!

i'm not knocking anyone here

It made me think though, as i have actually been visiting tomshardware since about 1997 but only signed up this year too. i could'nt possibly rack up even your number of posts let alone Jack's and wusy's as i work too many hours have have loads of commitments.

OP: interesting question. but i can't afford to try and find out, i have almost as big a battle trying to keep my rev c0 prescott cool
I can only wish that i knew 1/2 as much as Jack. :wink:
October 18, 2006 2:30:30 PM

At what temps does the E6600 fail at? I've yet to find a post of someone who burnt out their chip. My temps have run into the mid 70s before without any visible problem.
October 18, 2006 2:40:08 PM

Quote:
At what temps does the E6600 fail at? I've yet to find a post of someone who burnt out their chip. My temps have run into the mid 70s before without any visible problem.
Pumping 2.0v into one, with enough cooling that it won't throttle/shutdown would probably kill it in short order.
October 18, 2006 3:19:46 PM

Quote:
At what temps does the E6600 fail at? I've yet to find a post of someone who burnt out their chip. My temps have run into the mid 70s before without any visible problem.
Pumping 2.0v into one, with enough cooling that it won't throttle/shutdown would probably kill it in short order.

From what I understand, the biggest enemy to CPU's is not heat as much as it is voltage, though the two are related. Under air, you run into too much heat before you can apply enough voltage to fry it (in general), the converse is true with water, TEC, Phasechange. With those you can eaisly push too much voltage and still keep a reasonable level of heat. It is a combo deal.

I want to see my C2D in the 0-10C range when underload and fully OC'd. I am not entirely sure how realistic that is, but thats why they call these things experiments. Wusy is, of course, going way over my head with his knowledge and I very much want to hear the result.

My vision, one day in the next oooooo 5-10 years, is in my study in my house that I am going to build.... building in a LN cooling apparatus to the wall and have it in a closet of some nature with pipes leading out of the closet and in on one side of a built in desk, and on the ends of those pipes having some version of quick connectors, like on water hoses. The idea being that once the LN is set up, all I have to do is exchange the end pipes each time I rebuild my computer.

How's that for utterly ridiculous?

PS-might need to make friends with a chem student.... not like I know how to setup an entire LN assembly lol. :wink:
October 18, 2006 4:27:58 PM

If I had the money, no problem. I'd take that thing out as far as I could. Trouble is, the money. Taking out that far risks not only cpu, but the motherboard, graphics card, and anything else that gets in the way of flying electrons and frying circuits.

What? You smell smoke? Can't be, I've only got this thing up to 6 ghz. Gotta be another ghz left in it.
October 18, 2006 4:42:28 PM

So if the CPU will always cut out and prevent thermal damage, can I take it into the high 70s like it was before. It would then idle around 58, and only hit the high 70s on load. I could maybe push up the voltage to 1.45 or something.

I should be able to get my CPU to 3.6Ghz at those temps, but am really worried because I dont want to fry my 8 week old chip. That would be a horrid feeling.

Please advise.

PS. Wusy, ever since that OC'ing guide, you've developed GOD status in these forumz. Its hilarious. We done son.
October 18, 2006 4:50:34 PM

Quote:
From what I understand, the biggest enemy to CPU's is not heat as much as it is voltage, though the two are related. Under air, you run into too much heat before you can apply enough voltage to fry it (in general), the converse is true with water, TEC, Phasechange. With those you can eaisly push too much voltage and still keep a reasonable level of heat. It is a combo deal.

Bang on. For normal users it's about keeping temperature under throttling threshold. That's all, they just can't fail through overheating anymore.

To true overclockers, above ambient temp just isn't even taken into consideration, but rather how cold you can get it and sustain it long enough to pull off a benchmark or whatever one is trying to accomplish.
The rule of thumb is the lower temperaure you can achieve, the higher clock you can pull off with the same voltage to a certain extent.
You can never have enough cooling when overclocking.

For us going for Bungholiomarks, we're all using voltages that far exceeds point of radical eletron migration. So the aim is to find a voltage that will allow the CPU to attain maximum potential while still lasting over roughly a week before potential degrades. After that it's up to luck, tweaking and how low in temp you can get to get maximum Mhz.
A skilled overclocker will be able identify that ideal setting in only a short number of runs before fine-tuning to ultimate. That in turn increases number of meaningful benchmarks they can pull off.

Hence my quest for lower temperature.

So assuming I don't want to cause any shortened lifespan to my 6800 extreme, what is the maximum voltage that I could safely apply? are you saying that any overvoltage at all shortens life to some degree?
October 18, 2006 4:53:08 PM

And to my question to - what are the maximum temps I can run at. Are the high 70s ok?
October 18, 2006 5:00:22 PM

Quote:
So if the CPU will always cut out and prevent thermal damage, can I take it into the high 70s like it was before. It would then idle around 58, and only hit the high 70s on load. I could maybe push up the voltage to 1.45 or something.


Trouble is not only the heat, that smoke that rises from the computer, but those flying electrons. As Superfly pointed out, and Wusy confirmed, as voltages rise, burning starts taking place even if the cpu is kept cool. Electron migration will short the cpu out and you sit there wondering, "It never got hot. It was only 60c. Why did it stop?" Of course, while its shorting out the cpu, your motherboard can get in the way and do its own burn, and at the moment of the short, anything and everything that electrical and connceted can take a hit.

Relatively few have done such things, but I managed once years ago. When the repairman came out and started replacing parts (warrenty stuff, I claimed) I mentioned that it sparked only for a moment. He replyed that a moment is all it needed. I've been more careful since.
October 18, 2006 5:08:14 PM

And what about the temps. If electron migration may occur over 1.5V like Wusy said, at what temp do you start running the risk of losing performance/damaging your CPU??????

If it runs at 79C, and throttling is at 85C, will it be ok for me to do this, run at load at 79C and run idle at about 60?
October 18, 2006 5:11:31 PM

The typical rule of thumb is +/- 10% of the stock voltage is safe, I've been doing +15% on my Venice and it is still holding on. It is a matter of judgement and temperature. The colder you can keep it, the more you can stray outside the safe zone. At 70c I wouldn't want to stray too far outside the safe zone, because those temps are near throttling, aren't they? And as it has been pointed out, there is a max you can hit in voltage even with superior cooling.
October 18, 2006 5:15:31 PM

Apparently the chip throttles at 85, so I'm really trying to find out that if I max out at 79, if this would be ok. I dont want to gain 0.4Ghz at the expense of wrecking my chip, but am really now an addict to this world, and want to push it as far as I can. I have DDR2 800 4-4-4-15 RAM, and would really like to use my components to the max. If I can get 3.6Ghz, 400FSB X 9 stock, at the 1:1 multipler, then I think I will be set, cause thats about as high as I can go realistically.

3.6 Seems so far away at this stage. If I can push my temp up to 79 at load, I might be able to do it, BUT I DONT WANT TO SCREW UP MY CHIP.

Please advise. Thanks guys.
October 18, 2006 5:24:02 PM

Honestly, I am not sure how much more you can pull on that chip. You seem to be very close to the thermal limits. If there are no increases necessary in voltage, you might be able to pull it off, though those 70c+ numbers make me nervous a bit. The higher the chip is clocked the lower the temps need to be for it to be stable. If you are stable now, I would suggest not pushing it. I know, I am an addict of the OC world as well, and my bank account really proves that lol. 8O
October 18, 2006 5:35:31 PM

Thats really such a pity, but I will take your advice to heart and hold off. Im maxing at 70C now, and I think that this is as far as it goes. I havent event upped my voltage yet. I'm still using the auto function, and that puts it at 1.34 something or other. If I set it to 1.4, it jumps load temps by 5C, pushing the high 70s, and I only got an extra 10mhz out of that. I didnt try loosening my ram timings or play with other voltages or anything like that.

QUICK QUESTIONS - what the highest Ghz I can get on an E6600, running at 1.4V. Wuzy?
October 18, 2006 5:53:15 PM

Quote:
And what about the temps. If electron migration may occur over 1.5V like Wusy said, at what temp do you start running the risk of losing performance/damaging your CPU??????

If it runs at 79C, and throttling is at 85C, will it be ok for me to do this, run at load at 79C and run idle at about 60?


Used to be said that you want to keep the cpu as close to 40c as possible. Whether you choose very big air coolers or go water cooling or whatever, doesn't make a difference. Keep the chip cool. I personally think that 79c is way too high. Even if the cpu throttles itself at 85c for self protection, that doesn't mean that damage hasn't occured.

Oh yeah, besides cpu coolers, never underestimate the value of lots of fans to keep the case interior cool. I run 5 fans on mine at the moment, and am thinking of adding a 6th fan when I make a couple more upgrades. That's besides water cooling the video card. If the case interior gets hot and you're using air cooling, the cpu cooler will help, but not do as much good as it could if the case interior was cool.

Back to what Wusy said, that after 1.5V, electron migration starts. That voltage is partly independent of cpu temp, although a hotter cpu will allow migration faster than a cool cpu. You could have the cpu down at 40c and electron migration could start because of too high a voltage. Voltage that begins the destruction will vary among cpu's. Someone might get away with 1.55v, or even higher. The next will find his cpu burned at 1.45v. Its part of a game that's played in overclocking. If you can't afford a new cpu, stay within reasonable limits and you'll probably never have trouble.
October 18, 2006 7:14:50 PM

Quote:
QUICK QUESTIONS - what the highest Ghz I can get on an E6600, running at 1.4V. Wuzy?
I won't attempt to answer that because I don't have the sheer volume of knowledge to, but I can say it depends on just on your voltage to the CPU but also the chipset itself, as well as that temperature and the overall construction and configuration of your setup/mobo.

While it's possible to guess, as sailer pointed out, every chip is different, and with that so is every motherboard so a ball park figure is the best that can be accomplished. Even with a ball park, at the same voltage a CPu at 20C can achieve a higher OC than a 60c chip. So if you really want to push it to 3.5+ you are going to need to drop those temps honestly. Like sailer mentioned, those temps do make me nervous but the low 70's aren't unheard of, they are just unusual for C2D.
October 18, 2006 7:27:16 PM

I just remembered seeing you on the thread "Does cpu OCing see diminishing returns". JumpingJack had pushed a C2D up to very high overclocks. Got to keep in mind that he's a bit of a professional, that he knows what he is doing, and he knows that his cpu can and probably will burn at some point. Got to remember that playing with high overclocks, just like playing with fire, a person will get burned from time to time. If you can't afford disaster, don't play too hard.

I've got 3 computers. Two of them do not get played with, overclocking wise. They are business computers that I can afford to loose stuff on. One computer, the one I'm using at the moment, is the one I play with, both in games and in overclocking. I don't go for extremes, because I can't afford to replace hardware more than once a year or so. In the end, its all about knowing limits, whether limits with hardware, or limits with the pocketbook, or both.
October 18, 2006 7:28:17 PM

Quote:
Thats really such a pity, but I will take your advice to heart and hold off. Im maxing at 70C now, and I think that this is as far as it goes. I havent event upped my voltage yet. I'm still using the auto function, and that puts it at 1.34 something or other. If I set it to 1.4, it jumps load temps by 5C, pushing the high 70s, and I only got an extra 10mhz out of that. I didnt try loosening my ram timings or play with other voltages or anything like that.

QUICK QUESTIONS - what the highest Ghz I can get on an E6600, running at 1.4V. Wuzy?


i'm not wuzy, but perhaps this may help you frying up your hardware:
http://forumz.tomshardware.com/hardware/Core2Duo-Overclocking-Guide-ftopict197995.html

it's his great legacy anyway, enjoy! :) 
October 19, 2006 2:27:47 PM

Yeah, speculation is about all we can really do. We can say that operating a CPU w/o a heatsink is bad idea 8O

I mainly stay concerned about temperatures because there is an inverse relationship between temp and stable clock speed. The lower the temp the higher the stable clock speed, holding all other variables constant and assuming the CPU is the limiting factor.

Oh wusy, on the 7 series GeForce they are core speed limted so a 1MHz OC on the core is more helpful than a 1MHz OC on the memory... forgot to send you the results. Idk where they ran off to, MY Xp install is blow to hell, and I am scraping by on Vista RC2 until I get the time to properly reformat.
October 19, 2006 4:49:18 PM

Quote:
[Back to what Wusy said, that after 1.5V, electron migration starts.

I didn't say that...

My apologies. I was refering to what fishboi said that what you had said. That's what I get for answering someone else quote of what someone else said without seeing the actual quote myself. Again, sorry about that.

As to the 40c, my reference was to what "used to be said". I recognize that this value may longer be true, particularly with some given processors that are designed to run hotter than 40c, but I think excess heat is still an enemy that should be avoided.

I fully agree with the excess voltage problem. I've got a burned cpu to remind me of it.
!