Freezer?GC_PaNzerFIN wrote:
...Who uses his CPU 10 years 24/7?)
correct. I was about to add that.Sup wrote:
Freezer?GC_PaNzerFIN wrote:
...Who uses his CPU 10 years 24/7?)
3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8
That's absolute bollocks.GC_PaNzerFIN wrote:
Intel 65nm CPUs CAN take 70c just fine. In long term use too. My R600 load temps were over 90c it clearly wasn't dying. I know many guys with over 70c load temps for years with absolutely no probs. Quite many of them are oc guys that do know their job and they say it won't degrade at 70c. (maybe a week of its 10 years normal life time. who cares about that? Who uses his CPU 10 years 24/7?)
I think I believe the OC guys and official Intel statements...
The R600 is a GPU and GPUs can run way hotter than CPUs. 120C isn't too much of an issue for some GPUs. They use different fabrication processes for a start. Intel have released statements saying 60C is the upper threshold for safe operation on their 65nm chips.
And I thought you had some clue what you were talking about....
Comparing GPU and CPU temps - HA!
Last edited by Bertster7 (2008-05-03 08:09:13)
http://processorfinder.intel.com/detail … Spec=SLAPK
Thermal Specification: 72.4°C
And that's a 45nm... 73.2°C for the 2180.
Thermal Specification: 72.4°C
And that's a 45nm... 73.2°C for the 2180.
Last edited by Freezer7Pro (2008-05-03 08:12:30)
The idea of any hi-fi system is to reproduce the source material as faithfully as possible, and to deliberately add distortion to everything you hear (due to amplifier deficiencies) because it sounds 'nice' is simply not high fidelity. If that is what you want to hear then there is no problem with that, but by adding so much additional material (by way of harmonics and intermodulation) you have a tailored sound system, not a hi-fi. - Rod Elliot, ESP
+1Freezer7Pro wrote:
http://processorfinder.intel.com/details.aspx?sSpec=SLAPK
Thermal Specification: 72.4°C
And that's a 45nm... 73.2°C for the 2180.
bert... you were talking about silicon degrading at 60c. Thats why I posted what I did to show that is not the fact. and you said the imacs temps are 80c to 90c. THAT is too hot.
WUT?Bertster7 wrote:
And I thought you had some clue what you were talking about....
Last edited by GC_PaNzerFIN (2008-05-03 08:29:54)
3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8
GC_PaNzerFIN wrote:
Intel 65nm CPUs CAN take 70c just fine. In long term use too. My R600 load temps were over 90c it clearly wasn't dying. I know many guys with over 70c load temps for years with absolutely no probs. Quite many of them are oc guys that do know their job and they say it won't degrade at 70c. (maybe a week of its 10 years normal life time. who cares about that? Who uses his CPU 10 years 24/7?)
I think I believe the OC guys and official Intel statements...
GunslingerOIFII wrote: I read it on the Internet so it must be true - 90% of you guys.
i look at this thread and all i can wonder is:
what's the desktop?
what's the desktop?
"people in ny have a general idea of how to drive. one of the pedals goes forward the other one prevents you from dying"
Absolutely that's too hot.GC_PaNzerFIN wrote:
+1Freezer7Pro wrote:
http://processorfinder.intel.com/details.aspx?sSpec=SLAPK
Thermal Specification: 72.4°C
And that's a 45nm... 73.2°C for the 2180.
bert... you were talking about silicon degrading at 60c. Thats why I posted what I did to show that is not the fact. and you said the imacs temps are 80c to 90c. THAT is too hot.
As for those temperature specs, have you looked at the voltage ratings on them? No, I thought not.
1.5675V is much higher voltage than 1.3625V which is the top end of the operating margin for the E8500. That combined with the temperature is too much. That is critical temperature anyway, not the point at which it starts to be degraded, which is substantially lower.
How can you know guys who've run Conroe based CPUs at 70C for more than 3-4 years (which I think you'll find is the timeframe I mentioned in my original post) when they haven't been released for that long?
There are quite a few interesting articles out there with some nice graphs you can look at to see how much more likely your CPU is to fail at those sort of temps (65C+).Bertster7 wrote:
70C is pushing the boundaries of acceptable temps. It is too hot, but only just. It's fairly unlikely that the CPU won't last 3 or 4 years running like that. The problem is that if your ambient temps increase by 5-10C which in summer they often do, then your CPU becomes quite dangerously hot.
IC failure rate is a big deal. I had to do a course on it before.
Yup, apparently not.GC_PaNzerFIN wrote:
WUT?Bertster7 wrote:
And I thought you had some clue what you were talking about....
bert... the critical limit for C2D is 85c. That thermal spec is for 24/7 use...
edit: the throttle down temp is 85c and emergency shutdown is 90c.
edit: the throttle down temp is 85c and emergency shutdown is 90c.
Last edited by GC_PaNzerFIN (2008-05-03 09:30:54)
3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8
Time for a break. Cookie anyone?
My friend had over 75c idle temps and 84c load temps with his q6600 B3 rev. for 6 months. It still works just fine. He got better cooler now tho because I told him to get.
Also I have had E2180, E6320, E6750 and Q6600 G0 and I have seen what is the critical limit for each of them.
I had 1.525Vcore with my E6750 with load temps of 70c for 9 months. It didn't show any signs of degrading...
Also I have had E2180, E6320, E6750 and Q6600 G0 and I have seen what is the critical limit for each of them.
I had 1.525Vcore with my E6750 with load temps of 70c for 9 months. It didn't show any signs of degrading...
3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8
I don't wanna argue too much about stuff I don't know all too much about, but I think you're taking the numbers a bit too seriously, Bertster. You've got all the theory done, but I choose to believe Intel's numbers and my own experience with them over you.
Take for example my P4, which I've had for about three years. It's thermal spec is 67C. I oc'd it to 3.7GHz with 1.5VCore on the stock cooler, getting idle temps of 70+ and load temps of way over 80, almost 90C. I ran like that for about a year. During that year, the cooler fell off when I was at school, following a fan blade breaking. For six hours, the computer ran without even a heatsink. It didn't shut off, but I can assure you that, even throttling, it was running at over 100C.
That processor has survived another short heatsink failure running at 3.8, and almost two years sat 3.8GHz with 1.55VCore and load temps of 75C+. This processor is still not showing any signs of degradation, now running at 4GHz and 1.6VCore, temps ranging from mid-60's load to high 70's load, depending on the cooler I've had, which it has been doing since the 18th of January this year, 20 hours a day. If these numbers would be 100% correct and applicable, my CPU would have been dead and buried very long ago.
Take for example my P4, which I've had for about three years. It's thermal spec is 67C. I oc'd it to 3.7GHz with 1.5VCore on the stock cooler, getting idle temps of 70+ and load temps of way over 80, almost 90C. I ran like that for about a year. During that year, the cooler fell off when I was at school, following a fan blade breaking. For six hours, the computer ran without even a heatsink. It didn't shut off, but I can assure you that, even throttling, it was running at over 100C.
That processor has survived another short heatsink failure running at 3.8, and almost two years sat 3.8GHz with 1.55VCore and load temps of 75C+. This processor is still not showing any signs of degradation, now running at 4GHz and 1.6VCore, temps ranging from mid-60's load to high 70's load, depending on the cooler I've had, which it has been doing since the 18th of January this year, 20 hours a day. If these numbers would be 100% correct and applicable, my CPU would have been dead and buried very long ago.
The idea of any hi-fi system is to reproduce the source material as faithfully as possible, and to deliberately add distortion to everything you hear (due to amplifier deficiencies) because it sounds 'nice' is simply not high fidelity. If that is what you want to hear then there is no problem with that, but by adding so much additional material (by way of harmonics and intermodulation) you have a tailored sound system, not a hi-fi. - Rod Elliot, ESP
Freezer7Pro wrote:
During that year, the cooler fell off when I was at school, following a fan blade breaking. For six hours, the computer ran without even a heatsink. It didn't shut off, but I can assure you that, even throttling, it was running at over 100C.
16 minutes is hardly what I'd consider Orthos stable....
And that is very high voltage for 65nm... I wouldn't run it that way for long.
And that is very high voltage for 65nm... I wouldn't run it that way for long.
well :careface: about whether it degrades or not. If it does then he's got noone to blame but himself and it was only 50 bucks. If it doesn't then more power to him tbh.
This CPU is just there for the following month before he's away for miilitary. It doesn't need to last longCrazeD wrote:
16 minutes is hardly what I'd consider Orthos stable....
And that is very high voltage for 65nm... I wouldn't run it that way for long.
The idea of any hi-fi system is to reproduce the source material as faithfully as possible, and to deliberately add distortion to everything you hear (due to amplifier deficiencies) because it sounds 'nice' is simply not high fidelity. If that is what you want to hear then there is no problem with that, but by adding so much additional material (by way of harmonics and intermodulation) you have a tailored sound system, not a hi-fi. - Rod Elliot, ESP
You're talking absolute rubbish.GC_PaNzerFIN wrote:
bert... the critical limit for C2D is 85c. That thermal spec is for 24/7 use...
edit: the throttle down temp is 85c and emergency shutdown is 90c.
Those things are firmware controlled and totally external factors (yes the CPU will send a PROCHOT signal when it gets too hot, but it's up to the mainboard to decide what to with it). They have no reflection on the CPU itself. You're just demonstrating more and more you don't have a clue what you're talking about and you're just using a few examples, badly chosen examples, to illustrate a stupid point.
Everyone knows that all CPUs are different and that the chances of being able to run one beyond temps that other ones will fail at are there. That does not prevent those temps from being too high.
You've got to be a really retarded overclocker to believe that erring on the side of caution, rather than the absolute limits of what CPUs will take, is a bad idea. 70C is too hot. It's that simple. The critical limit for a C2D is nowhere near 85C. The thermal spec for an E6300 for example, is 61.4C, for an E6600 it is 60.1C, specs from Intel (I don't understand why you keep saying you believe Intel over me, when all Intel's figures agree with what I'm saying absolutely). Of course for the E2180 the max temps are somewhat higher because of the massively reduced cache density.
Here is a good guide, based on facts and not gut experience based on a few sample cases, on maximum OCing temps.
You are just being wrong in the most stupid and dangerous way. Advising people to take overclocks to the absolute limit and run them at those clocks for sustained periods is not only ludicrous, but potentially dangerous for other peoples expensive hardware - something I have to admit I take exception to.
That's a P4. Totally different manufacturing process. Totally different type of silicon. Very different scale and density. These factors all equate to a totally different scenario. Which is not to say that a Conroe couldn't survive that, it's just less likely (to be honest it's quite lucky that your P4 survived that).Freezer7Pro wrote:
I don't wanna argue too much about stuff I don't know all too much about, but I think you're taking the numbers a bit too seriously, Bertster. You've got all the theory done, but I choose to believe Intel's numbers and my own experience with them over you.
Take for example my P4, which I've had for about three years. It's thermal spec is 67C. I oc'd it to 3.7GHz with 1.5VCore on the stock cooler, getting idle temps of 70+ and load temps of way over 80, almost 90C. I ran like that for about a year. During that year, the cooler fell off when I was at school, following a fan blade breaking. For six hours, the computer ran without even a heatsink. It didn't shut off, but I can assure you that, even throttling, it was running at over 100C.
And again bringing Intel's numbers, which support what I'm saying absolutely, into it as an argument against the point I'm making. If you don't understand the numbers Intel publishes, that's all the more reason to be more careful with your overclocks.
That voltage is very high. Dangerously outside manufacturer tolerances. Also, I hadn't noticed it was only 16mins...CrazeD wrote:
16 minutes is hardly what I'd consider Orthos stable....
And that is very high voltage for 65nm... I wouldn't run it that way for long.
Load testing takes at least 3-4 hours. Any less than that gives no proper indication of how hot the thing will get or how stable it will continue to run.
Last edited by Bertster7 (2008-05-04 05:34:50)
What exactly does the E6600 numbers have to do with this? Yes, they're lower, but we're talking about an E2180 here. I took my P4 as an example that processors can survive more than it says on the paper. I know it's all different than a Conroe, but it's thermal spec is lower than many Conroes, especially the 2180, and it's been running a lot hotter than any Conroe for a pretty long time.Bertster7 wrote:
That's a P4. Totally different manufacturing process. Totally different type of silicon. Very different scale and density. These factors all equate to a totally different scenario. Which is not to say that a Conroe couldn't survive that, it's just less likely (to be honest it's quite lucky that your P4 survived that).Freezer7Pro wrote:
I don't wanna argue too much about stuff I don't know all too much about, but I think you're taking the numbers a bit too seriously, Bertster. You've got all the theory done, but I choose to believe Intel's numbers and my own experience with them over you.
Take for example my P4, which I've had for about three years. It's thermal spec is 67C. I oc'd it to 3.7GHz with 1.5VCore on the stock cooler, getting idle temps of 70+ and load temps of way over 80, almost 90C. I ran like that for about a year. During that year, the cooler fell off when I was at school, following a fan blade breaking. For six hours, the computer ran without even a heatsink. It didn't shut off, but I can assure you that, even throttling, it was running at over 100C.
And again bringing Intel's numbers, which support what I'm saying absolutely, into it as an argument against the point I'm making. If you don't understand the numbers Intel publishes, that's all the more reason to be more careful with your overclocks.
The idea of any hi-fi system is to reproduce the source material as faithfully as possible, and to deliberately add distortion to everything you hear (due to amplifier deficiencies) because it sounds 'nice' is simply not high fidelity. If that is what you want to hear then there is no problem with that, but by adding so much additional material (by way of harmonics and intermodulation) you have a tailored sound system, not a hi-fi. - Rod Elliot, ESP
It ran orthos over an hour before I did a reboot myself... I had the CPU installed for 20min at that time...
And Bertster... Do you have any own experience with C2Ds/C2Qs at over 70c? Sure you know the theoryt but real life isn't like that when it comes to CPU lifetime...
And Bertster... Do you have any own experience with C2Ds/C2Qs at over 70c? Sure you know the theoryt but real life isn't like that when it comes to CPU lifetime...
Last edited by GC_PaNzerFIN (2008-05-04 05:42:36)
3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8
The E6600 is a C2D.Freezer7Pro wrote:
What exactly does the E6600 numbers have to do with this? Yes, they're lower, but we're talking about an E2180 here. I took my P4 as an example that processors can survive more than it says on the paper. I know it's all different than a Conroe, but it's thermal spec is lower than many Conroes, especially the 2180, and it's been running a lot hotter than any Conroe for a pretty long time.Bertster7 wrote:
That's a P4. Totally different manufacturing process. Totally different type of silicon. Very different scale and density. These factors all equate to a totally different scenario. Which is not to say that a Conroe couldn't survive that, it's just less likely (to be honest it's quite lucky that your P4 survived that).Freezer7Pro wrote:
I don't wanna argue too much about stuff I don't know all too much about, but I think you're taking the numbers a bit too seriously, Bertster. You've got all the theory done, but I choose to believe Intel's numbers and my own experience with them over you.
Take for example my P4, which I've had for about three years. It's thermal spec is 67C. I oc'd it to 3.7GHz with 1.5VCore on the stock cooler, getting idle temps of 70+ and load temps of way over 80, almost 90C. I ran like that for about a year. During that year, the cooler fell off when I was at school, following a fan blade breaking. For six hours, the computer ran without even a heatsink. It didn't shut off, but I can assure you that, even throttling, it was running at over 100C.
And again bringing Intel's numbers, which support what I'm saying absolutely, into it as an argument against the point I'm making. If you don't understand the numbers Intel publishes, that's all the more reason to be more careful with your overclocks.
I was illustrating the point that different CPUs within the same series have differing thermal characteristics.GC_PaNzerFIN wrote:
the critical limit for C2D is 85c. That thermal spec is for 24/7 use...
You are still clinging to isolated examples, which in a world of averages are utterly meaningless. 70C is too hot for sustained usage. Simple as that. If you want to take risks with your CPU and do risky on the edge OCing, then that's fine - but don't recommend it to others and don't claim it's fine. It can lead other people to base what they consider to be safe temps on the gibberish you guys have been spouting.
70C is not a safe temp. It is too hot (particularly when only tested for 16mins (that's a joke) and at such extreme voltages which put the chip way past max. TDP).
Yes. Lots of experience with it.GC_PaNzerFIN wrote:
It ran orthos over an hour before I did a reboot myself... I had the CPU installed for 20min at that time...
And Bertster... Do you have any own experience with C2Ds/C2Qs at over 70c? Sure you know the theoryt but real life isn't like that when it comes to CPU lifetime...
I see about 5 different machines fitting those criteria every day. Typically about 1 in 6 will fail.
It is all very well to OC and run CPUs hot. But you have to be aware that there is a definite element of risk involved when running at those sort of temps. Claiming that 70C is a safe temp to be running at when using so much voltage is just not true and could lead other less experienced people to damage lots of expensive hardware.
Last edited by Bertster7 (2008-05-04 05:47:04)
WTF is wrong with you?!?!!? I will go 1.6vcore just to show you it doesn't blow up... you can run C2Ds at 70c just fine. I asked guy working for Foxconn as Intel CPU overclocker and he said 70c is fine...Bertster7 wrote:
70C is not a safe temp. It is too hot (particularly when only tested for 16mins (that's a joke) and at such extreme voltages which put the chip way past max. TDP).
edit: the iMACs running at OVER 80c is a lot different thing...
Last edited by GC_PaNzerFIN (2008-05-04 05:46:21)
3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8
He's wrong. Sustained anyway. Fine for short term usage. I haven't said it's going to blow up, and the fact you are thinking in those terms shows even more you don't have a clue. Junction temps in excess of 70C are damaging. They create a definite possibility of sudden, random gate failure and certainly reduce the lifespan of the CPU to a fraction of what it should be.GC_PaNzerFIN wrote:
WTF is wrong with you?!?!!? I will go 1.6vcore just to show you it doesn't blow up... you can run C2Ds at 70c just fine. I asked guy working for Foxconn as Intel CPU overclocker and he said 70c is fine...Bertster7 wrote:
70C is not a safe temp. It is too hot (particularly when only tested for 16mins (that's a joke) and at such extreme voltages which put the chip way past max. TDP).
edit: the iMACs running at OVER 80c is a lot different thing...
All the technical documents and all my (and my colleagues, some of whom currently work for Intel) extensive experience with this reach exactly the same conclusion. If you'd like to produce some evidence to back up your silly outlandish claims, you are welcome to do so. But you won't find a technical document anyway that says you're right.
You think the guys working at Intel think their thermal specs are jokes?
3930K | H100i | RIVF | 16GB DDR3 | GTX 480 | AX750 | 800D | 512GB SSD | 3TB HDD | Xonar DX | W8