E6600 running very hot

E

ex0du5

Gawd
Joined
Jun 26, 2007
Messages
873
Ok well I have an E6600, P5B-E, and a Zalman 7700 fan/heatsink.

The fan is mounted onto some special Zalman bracket, as the fan design is not meant to screw into the standard 4 corner threads on my P5B-E. Also, I have the heatsink seated with some AS5 (first I tried the line method, then I just put some more on to see if it would improve). I even made sure to screw the heatsink in as tight as I possible could (even stripped the screw head a bit :().

Well, CoreTemp, and RM Clock, both report my core temp to be 45degrees IDLE stock @ 2.4GHz. This is absurdley high! I reach about 59-60 running TAT on 100% at these stock speeds. There's no way I can overclock without pushing the temps much higher than I should have to.

Why on earth are my temps so high? My room temperature is roughly 21-22C
 
I'd switch to a better cooling unit. The Zalman 7700 isn't that good by today's standards. BTW as long as your load temperature doesn't exceed 65c you won't be hurting anything.

I don't like the way the Zalman units mount to motherboards and I never have. My advice is get the Thermalright Ultra Extreme 120. That's a badass cooler and it will serve you well. It is expensive, but it keeps my Pentium Extreme Edition 965 decently cool and that's saying alot.
 
wow... I don't mean to be a dick, but I've been at the [H] for over 6 months, and literally EVERY day there is a post about "why does my Core2 idle so hot??" Dear God, take a look at the overclocking page or Intel page, I'm sure you'll find 3 posts on the page about your idle temp, most likely with your same degrees. I don't even own a Core2 but I can tell you that:
1. At lower temps the core temps aren't accurate, as much as 8-10 degrees +/-, so don't worry about it.
2. Why does idle matter so much anyways? It's what your load temps are that should worry you. Run Orthos and see how hot it gets, if THAT is too hot then you have a problem. If not, then you're fine. 30-40C isn't THAT hot for a processor!
 
Hes not just complaning his chip is to hot. He's asking why its running so hot (if his heatsink is mounted wrong) or if its normal. And those temps might be fine for a stock cooler, but any aftermarket heatsink should drop it down to atleast 35C idle at stock speeds and hes already stated that he wants to overclock so obviously he needs the lowest temps possible.

I'd dump that heatsink, and like stated grab a Thermalright Ultra Extreme 120, one of the best air cooling units. If your partial to zalman though the zalman CNPS9700 works very well (its what I use) but for that price id go with the thermalright anyways.
 
Well even though those temps are still good as in its not going to hurt your processor wise, something is defentely wrong there. Check to make sure that processor is making full contact with the processor. I know the LGA 775 bracket you are talking about, and even though I dont have experience with it and that 7700, it could be possible you got a defect, or maybe attached it in some wrong way. I would make sure you check everything.

I got the new Scythe Ninja Rev B where it uses a new mounting bracket cause the 680i's have some capicator on the backside of the motherboard and it wasnt making full contact with the processor. I found out the bracket was defected and just returned the whole thing and got a Ultra 120.
 
You may have too little thermal paste or not enough. I've run into this issue a couple of times. It really can make a huge difference in your temperatures.
 
Dan_D said:
You may have too little thermal paste or not enough. I've run into this issue a couple of times. It really can make a huge difference in your temperatures.
Click to expand...

Too little or too much? either or is not good ;)
 
With my apogee, and overclocked to 1875 FSB (E6300) with 1.5 (1.45), I get 40C idle and 55C load, is that good? Do i need better cooling? When everything is stock, I get 30C idle and about 40-42C load.
 
DangerIsGo said:
With my apogee, and overclocked to 1875 FSB (E6300) with 1.5 (1.45), I get 40C idle and 55C load, is that good? Do i need better cooling? When everything is stock, I get 30C idle and about 40-42C load.
Click to expand...

I have the apogeegt on a E6600@3400 1,46v and hits 47 load when dual Priming.
 
I think I'm just going to stick the stock Intel cooler on there. All I want for now is to run it at 3GHz, and with the Zalman, my load temps are getting to about 70 under TAT.

I'll see how the stock cooler does tonight. If there's something wrong there, then my processor must be extremely convex!
 
ex0du5 said:
I think I'm just going to stick the stock Intel cooler on there. All I want for now is to run it at 3GHz, and with the Zalman, my load temps are getting to about 70 under TAT.

I'll see how the stock cooler does tonight. If there's something wrong there, then my processor must be extremely convex!
Click to expand...

Sounds like it's time to lap ;)
 
Citizen86 said:
wow... I don't mean to be a dick, but I've been at the [H] for over 6 months, and literally EVERY day there is a post about "why does my Core2 idle so hot??" Dear God, take a look at the overclocking page or Intel page, I'm sure you'll find 3 posts on the page about your idle temp, most likely with your same degrees. I don't even own a Core2 but I can tell you that:
1. At lower temps the core temps aren't accurate, as much as 8-10 degrees +/-, so don't worry about it.
2. Why does idle matter so much anyways? It's what your load temps are that should worry you. Run Orthos and see how hot it gets, if THAT is too hot then you have a problem. If not, then you're fine. 30-40C isn't THAT hot for a processor!
Click to expand...

Well you certainly came off as a dick! You didn't even read my full post before flaming me! Take a look and see how I mention that I'm running 59-60 under load with an aftermarket cooler! In what way dose that not seem hot?

Citizen86 said:
Sounds like it's time to lap ;)
Click to expand...

Yeah :). If the Intel doesn't work out, then I might consider lapping.
 
DangerIsGo said:
With my apogee, and overclocked to 1875 FSB (E6300) with 1.5 (1.45), I get 40C idle and 55C load, is that good? Do i need better cooling? When everything is stock, I get 30C idle and about 40-42C load.
Click to expand...

Well, I was water cooling my E6600 briefly, but I don't remember what my temps were at 3.6GHz. My Core 2 Quad on the other hand idles at 43c and gets to about 52c under load at speeds up to 3.4GHz.
 
60C while running TAT is normal considering your cooler is only average.

TAT was designed to bake laptops and isn't a reasonable stability test because there isn't one real world piece of software that will ever heat your processor up the way TAT does. Even Orthos is a little over kill these days. Being able to run that for 4 hours is sort of the [H]ard standard to get into the OC database but I've yet to find a program that creates as much heat as Orthos does let alone TAT.

The other problem is that TAT does not exactly track the on chip digital thermal sensors ( DTS ) the way it should. People trust it because it says Intel on it but using CrystalCPUID to read the sensor directly will show you that from day to day it varies for no particular reason that I can come up with.

Your MHz is not nearly as important as your core voltage under load and you didn't post that information. Leaving it in the bios at AUTO can result in pretty much anything. Everest or SpeedFan will report this correctly and so will CPUz as long as you're under 1.40 volts.

Idle temperatures based on the digital thermal sensors that TAT and CoreTemp use to report temperatures may not be accurate. When idle temperatures can be reported that are less than the ambient temperature then something is definitely wrong. Until someone comes up with an explanation for this then your guess is as good as mine. Read this post for more info.

http://www.hardforum.com/showpost.php?p=1031184411&postcount=10
 
Dan_D said:
Too little or too much are bad. Though I find too little to be a worse situation than too much, though too much is messy.
Click to expand...

I agree, and AS5 is a pain to get off your fingers... I've put on too much before, had it seeping out a bit from the side, went to feel the base of my heatsink to see how hot it was.... ewwww.
 
Haha yeah I ended up getting some under my forearm without noticing and rubbed it all over my computer desk. Twas a pain to clean up.

Oh and my VCore is 1.28 at stock speeds (set to Auto).

Well at 3GHz, TAT runs my cores up to about 70C. Should I consider that adequate cooling for now? I don't plan on going past 3GHz just to keep things simple, and to avoid overclocking my memory.
 
AS5 will find a way into your food if you don't watch it.
 
Get the Zalman 9700 is the best air cooler. I have QX6700 idle 54c in core temp and load gets around 72c. It was much higher with the intel crappy stock cooler. The quadcores run a lot hotter then the Conroe or the core 2 dual. So their temp spec is much higher.
I have my CPU over clock to 3.2ghz from 2.66ghz default.
 
∞Velocitymaster∞;1031195808 said:
Get the Zalman 9700 is the best air cooler. I have QX6700 idle 54c in core temp and load gets around 72c. It was much higher with the intel crappy stock cooler. The quadcores run a lot hotter then the Conroe or the core 2 dual. So their temp spec is much higher.
I have my CPU over clock to 3.2ghz from 2.66ghz default.
Click to expand...

What? The Zalman CNPS9700 is far from the best cooler out there. There are several better coolers that come to mind like the Tuniq Tower and the Thermalright Ultra 120 Extreme. 54c idle and 72c load isn't good considering the safe limit for these CPUs is 65c.
 
Dan_D said:
72c load isn't good considering the safe limit for these CPUs is 65c.
Click to expand...
That 65C number is news to me. Do you have any documentation from Intel to support that?

You can reliably run Orthos 24/7 a lot higher than that if you're not overclocking and Intel will stand behind you 100%.
 
Dan_D said:
Read the specifications for yourself.
Click to expand...
That's the problem. I have read the Intel Datasheet and there is nowhere in there that mentions the 65C number that you quoted.

It's obviously in there somewhere or else you wouldn't have mentioned it. You posted a link to a 118 page document so could you be a little more specific than that and give us a page number so we can all read it.

I did find this in there:

"As an output, PROCHOT# (Processor Hot) will go active when the processor temperature monitoring sensor detects that

the processor has reached its maximum safe operating temperature.

This indicates that the processor Thermal Control Circuit (TCC) has been activated, if enabled. As an input, assertion of PROCHOT# by the system will activate the TCC, if enabled. The TCC will remain active until the system de-asserts PROCHOT#. See Section 5.2.4 for more details."

Processor Hot ( PROCHOT# ) is not asserted until DTS=0 which is an absolute core temperature of 85C on the Core 2 Duo processors, minimum. On Core processors that have a TjMax = 100C like the Quad processors, it is perfectly safe to operate your processor up to a core temperature of 100C. Intel clearly says that temperature is the "maximum safe operating temperature."

Edit: Anyone interested in learning more about Core temperatures should read this document:
http://irevues.inist.fr/bitstream/2042/6579/1/TMI23.pdf

Here's a nice quote:
"Functionality, electrical specifications and reliability commitments are guaranteed at maximum Tj as measured by the DTS."

That quote is at the end of the document just before summary and conclusions.
 
Section 5.1.1 of the document contains this data. It reads "To allow for the optimal operation and long-term reliability of Intel processor-based systems, the system/processor thermal solution should be designed such that the processor remains within the minimum and maximum case temperature (tc) specifications when operating at or below the Thermal Design Power (TDP) value listed per frequency in Table 25."

Further more the table lists the QX6700 as a 130watt part and the Q6600 as a 105watt part. Therefore refering to the tables referenced by Table 25 you can see the thermal profile for 130watt processors listed. It shows the MAXIMUM temperature of the QX6700 at full power (130watts) is 64.5c. Tables 14 and 15 also show similar data.

In short, 65c is the safe thermal design limitation for the processor(s) in question. The Q6600 is actually rated for only 62.2c. Each processor needs to also be exposed to temperatures below 5c according to the tables.

So again, I am not making this stuff up. I am not saying you can't successfuly run your processor at higher temperatures, but if Intel states that it's outside the design specifications for the chip, you probably shouldn't be going outside of those specs should you desire you CPU to live any length of time.
 
Dan_D: The problem is that you are trying to compare a case temperature number to a core temperature reading. CoreTemp, TAT and SpeedFan report core temperatures. These are two different temperatures, measured differently and can't be compared.

The case temperature is measured on top of the heat spreader at the geometric center. This location is on the edge between the two cores and the on chip L2 cache memory on the Core 2 Duo. A core temperature reading is the maximum value obtained after comparing multiple digital heat sensors that have been calibrated and strategically integrated into the core by Intel at all of the known hot spots.

Intel went to the digital thermal sensors because the old way of taking one reading and then trying to guess at what the rest of the processor was doing was not allowing users to get the most out of their processor. The new DTS method cured this problem.

Intel clearly states that the "maximum safe operating temperature" for a Core 2 Duo is reached when the digital thermal sensor ( DTS ) reports zero. For my E6400 that occurs at a core temperature of 85C. Running at this temperature is perfectly safe according to Intel and you will not damage your processor or negatively effect your warranty if you do so.
 
unclewebb said:
Dan_D: The problem is that you are trying to compare a case temperature number to a core temperature reading. CoreTemp, TAT and SpeedFan report core temperatures. These are two different temperatures, measured differently and can't be compared.

The case temperature is measured on top of the heat spreader at the geometric center. This location is on the edge between the two cores and the on chip L2 cache memory on the Core 2 Duo. A core temperature reading is the maximum value obtained after comparing multiple digital heat sensors that have been calibrated and strategically integrated into the core by Intel at all of the known hot spots.

Intel went to the digital thermal sensors because the old way of taking one reading and then trying to guess at what the rest of the processor was doing was not allowing users to get the most out of their processor. The new DTS method cured this problem.

Intel clearly states that the "maximum safe operating temperature" for a Core 2 Duo is reached when the digital thermal sensor ( DTS ) reports zero. For my E6400 that occurs at a core temperature of 85C. Running at this temperature is perfectly safe according to Intel and you will not damage your processor or negatively effect your warranty if you do so.
Click to expand...

If you have your processor running at 85c, do you really think for an instant that the case temperature is NOT beyond specification? I'd be willing to wager that if your internal core temperature is 85c, then the case temperature is going to be over 65c.

I'm looking for more data on the maximum core temperatures, and when I find it I would bet that I'm right and 75c+ will be outside of specifications. I could be wrong, but when you start talking about 85c then your talking about Geforce 8800GTX heat ranges which seems unlikely to me given that the Core 2 Quad has fewer transistors and pulls less power than a G80 core does.
 
If you think a core temperature of 85C is hot then have a look at the Core 2 Duo mobile processors. Each and every one of them on the Intel website is rated to be 100% safe and reliable up to a core temperature of 100C.

From the T2300E to the T7700, they all have the same maximum safe core temperature of 100C. Even the low watt processors like the T2300E with a thermal design power of only 31 watts still carries a 100C rating.

There's nothing special about the mobile processors. They are built in the same factories and some of them like the T7600 have the exact same 0x6F6 CPUID that the revision B2 desktop E6300, E6320, E6400, E6420, E6600 and E6700 have. They have a lower watt rating because they have been tested by Intel to be able to run reliably at a lower core voltage but other than that there isn't much difference, if any, between the two.

Here's the thermal design guidelines for the QX6800 quad with a 130 watt TDP rating:
http://processorfinder.intel.com/details.aspx?sSpec=SL9UK

Does that mean the quads have to stay below a core temperature of 54.8C? I don't think so.
 
unclewebb said:
If you think a core temperature of 85C is hot then have a look at the Core 2 Duo mobile processors. Each and every one of them on the Intel website is rated to be 100% safe and reliable up to a core temperature of 100C.

From the T2300E to the T7700, they all have the same maximum safe core temperature of 100C. Even the low watt processors like the T2300E with a thermal design power of only 31 watts still carries a 100C rating.

There's nothing special about the mobile processors. They are built in the same factories and some of them like the T7600 have the exact same 0x6F6 CPUID that the revision B2 desktop E6300, E6320, E6400, E6420, E6600 and E6700 have. They have a lower watt rating because they have been tested by Intel to be able to run reliably at a lower core voltage but other than that there isn't much difference, if any, between the two.

Here's the thermal design guidelines for the QX6800 quad with a 130 watt TDP rating:
http://processorfinder.intel.com/details.aspx?sSpec=SL9UK

Does that mean the quads have to stay below a core temperature of 54.8C? I don't think so.
Click to expand...

I see where you are coming from, but I still wouldn't feel comfortable having my Core 2 Quad run at 85c.

If the white papers for the desktop parts say the thermal solution needs to keep the temperature of the processor casing down to sub-65c temperatures, I have to believe that a core temperature of 100c isn't safe. You may believe the mobile parts to be the same, and at the core level they are damn close but the whole package that they are in is completely different. For one thing they don't have heat spreaders and for another thing the extra thermal tolerance may be exactly why the mobile parts are more expensive than the desktop parts. That and their ability to run on less power. So while I understand your reasoning here, I can't help but think it's somehow flawed.
 
Many people have these preconceived fears about running their cpu at a temperature that they are uncomfortable with. That's their fear but it's not based on fact.

During testing my E6400 does not start to throttle until DTS = 2 which is equivalent to a core temperature of 83C. Intel is not dumb. They're well aware that there are going to be plenty of users overclocking and over volting their processors, many with no common sense. They don't want to end up with a bunch of RMA claims because their processors couldn't handle the heat. That's why they designed the DTS. They want users to be able to safely run their processors as hard as possible without hurting them.

Thermal throttling and thermal shut down is all based on what the DTS is reporting. If a processor could not safely be operated beyond a core temperature of 65C then Intel would have set the throttle and shut down points 20C lower but they didn't. Not only can you reliably run your processor up to 85C but I think there is still another 20C of head room before the processor will shut itself down to prevent a melt down.

I once hit an estimated 97C on core0 while it was throttling. It just kept running and I never did hit the thermal trip point where the processor shuts itself down. The only thing that stopped me were my own fears. When going beyond 85C on an E6400 it is officially running out of spec and its output can no longer be guaranteed, so I quit testing.

Here's a couple of pictures I posted on XS. The only way I could create some serious heat was to remove the heatsink.
testbenchsg0.jpg


It started throttling right where it was supposed to:
84uh0.jpg


And after the torture test I ran it real hard and got into the [H]ard OC database not too far away from the quickest air cooled E6400 processors.

This is the best Intel document that explains TDP.
http://download.intel.com/design/processor/designex/31368504.pdf

Here's a quote from Section 4.2.6

"The processor TDP is based on measurements of processor power consumption while running various high power applications. This data is used to determine those
applications that are interesting from a power perspective. These applications are then evaluated in a controlled thermal environment to determine their sensitivity to
activation of the thermal control circuit. This data is used to derive the TDP targets published in the processor datasheet.

A system designed to meet the thermal profile at TDP and TC-MAX values published in the processor datasheet greatly reduces the probability of real applications causing the thermal control circuit to activate under normal operating conditions. Systems that do not meet these specifications could be subject to more frequent activation of the thermal control circuit depending upon ambient air temperature and application power profile. Moreover, if a system is significantly under designed, there is a risk that the Thermal Monitor feature will not be capable of maintaining a safe operating temperature and the processor could shutdown and signal THERMTRIP#."

The TDP numbers and charts are only guidelines for system builders to build systems that will keep away from throttling and from shutting down due to overheating of the processor. The case temperature guidelines are for system builders that use their own heatsinks. Intel has already tested their OEM heatsinks and fans and knows that their boxed processors will allow their processors to run within spec when used in a properly ventilated case. System builders that use the Intel boxed cooler can also use these guidelines when designing a case and deciding on how many case fans they should be using.

The case temperature has nothing to do with the throttling point, the shut down point or what temperature a Core 2 Duo can run 100% reliably and safely at.
 
I've had all sorts of fun learning this stuff thanks to all of your help in the forums.

currently i'm running a B2 Stepping E6600 with Arctic Cooler Pro 7
gigabyte P35-DS3R + 2GB Gskill 6400
I'm pushing high voltage (1.50 after VDROOP, 1.56 in BIOS)
running 9 x 379 @ 3.41 GHZ

I don't use TAT any longer for stressing, but ORTHOS, and at full load for 25 min, my temps (using both everest, coretemp .95, and TAT) are reading max at 60C hovering around 57 in a room which is 23C.

I'm happy with this setup, considering the high voltage and relatively manageable temps!!!

Thanks to all of you who have led me through this process!!

Its funny, I mod my WRX, Computer, Cameras, etc - - - -

why can't we just leave things alone??!!! hahah :p
 
unclewebb said:
Edited because that post was huge. -Dan
Click to expand...

I understand what your saying, and I get your point. I just think that 85c is unreasonable. Hell even on air my Core 2 Quad Q6600 doesn't run that hot overclocked to any setting it can handle. It's not like my ambient temperatures are that great, and I wasn't using the best air cooler either. Makes me wonder what the hell some of you are doing to get temps that high.
 
Very interesting read unclewebb.

I guess I'm still in the camp of being afraid to run much higher than 65C, just for losing some life of the processor. The thing is, if I run it at 70C constantly, I don't know how much life it will lose. I'll probably be keeping this processor at least another year or two, and give it to someone after, so I don't want to kill it off much faster than I need to.

I'm sure that with more experience running at these temps, I'll have a better idea in the future of where I want to keep my temps.
 
ex0du5 said:
Very interesting read unclewebb.

I guess I'm still in the camp of being afraid to run much higher than 65C, just for losing some life of the processor. The thing is, if I run it at 70C constantly, I don't know how much life it will lose. I'll probably be keeping this processor at least another year or two, and give it to someone after, so I don't want to kill it off much faster than I need to.

I'm sure that with more experience running at these temps, I'll have a better idea in the future of where I want to keep my temps.
Click to expand...

As long as your processor isn't throttling down your temps are probably ok. I don't know how much life you'd loose, but I suspect it isn't much. Even so I don't think 85c for a Core 2 Quad is ok. My Pentium D doesn't get that hot at 4GHz on air cooling.
 
This may or may not help. I have 2X E6600's, one is water cooled and the other is air cooled by a Zalman 9700 HSF. The air cooler is in a Antec 900 case. I like the "bling" of the Zalman HSF even if it's not the best. My ambient temp is by a indoor/outdoor thermometer and I fold 24/7. My temp ATM is 36c and the ambient temp ATM is 21c. My temps are read by Core Temp (hottest core) and monitored constantly by SpeedFan. I know the software programs may be inaccurate so I only use them for reference. I always use the "stripe" method of Artic Silver 5 and I always use Artic Silver 5 TIM, I find it the best overall. I have tried the ZM-STG1 (good for chipsets in my book) and Shin Etsu TIM, but I like AS5 best. I have a E6420 (OC'ed to 2.4 GHz) cooled by a TT Big Typhoon (AS5 TIM) and it's temp is ATM 35c and I have a E6400 (stock clock) cooled by a Zalman 9700 (also AS5) and it's temp by TAT is 34c (I don't know what their wattage is, probably lower than the E6600's 65w)

Edit: Oh yeah, I would have posted screen shots, but I don't know how. Would someone please explain how for this "dumby"?

Edit again: I just reread this post. I also have a Apoogee GT WB and folding 24/7 it gets about 29c to 30c with a ambient temp of 21c. Maybe the temp readings are screwed up ?
 
Dan_D said:
Even so I don't think 85c for a Core 2 Quad is ok. My Pentium D doesn't get that hot at 4GHz on air cooling.
Click to expand...
That's another problem. People that have been around for a while try to make temperature comparisons to the previous Pentium 4 line. The Core processors are the first ones that let a user read the hottest temperature on the core in real time. Once again, you can't directly compare core temperatures with case or diode temperatures.

The core temperature is going to be a higher number and definitely a more accurate number than the temperature number that is reported from a single diode in the Pentium 4 era.

This greater accuracy has allowed Intel to set the Core 2 Duo throttling and shut down points sky high. With a single diode they were previously forced to leave a lot more head room before the throttling / shut down point.

The digital thermal sensors have changed all that. No matter what type of software you're running and no matter what part of the core you're stressing, the DTS compares the temperatures of all of the known hot spots on the chip and its output is not the average temperature but the temperature for the hottest point. If any area of the cpu is out of spec then the entire processor is throttled back in MHz by first dropping the multiplier to 6X and it will also try to reduce the core voltage if the motherboard is willing. If the temperature of the processor can still not be controlled then sooner or later it will simply shut down.

All my testing can be summed up quite simply. People worry way too much about the temperature of their processors when there is absolutely no reason to. If you're not overclocked then it will likely take a CPU fan failure before you'll ever get up to the throttling or shut down point.

If you're overclocking then you are much more likely to lose stability before you ever reach the throttling point. Damaging a core processor by overheating it is very unlikely. Intel has put 110% into the thermal management of these processors and they have been specifically designed to shut down before any heat related damage can occur.

The other big concern is long term damage or life of the processor. The boxed Intel processors have a 3 year warranty. Three years from now processors will definitely be cheaper and you'll most likely be richer so $100 for a new one shouldn't be a major concern. I still have a well overclocked Pentium III running strong many years later. It also lived through a stupid user pulling the heatsink and fan off because he was curious :rolleyes: with no long term negative effects.
 
unclewebb said:
That's another problem. People that have been around for a while try to make temperature comparisons to the previous Pentium 4 line. The Core processors are the first ones that let a user read the hottest temperature on the core in real time. Once again, you can't directly compare core temperatures with case or diode temperatures.

The core temperature is going to be a higher number and definitely a more accurate number than the temperature number that is reported from a single diode in the Pentium 4 era.

This greater accuracy has allowed Intel to set the Core 2 Duo throttling and shut down points sky high. With a single diode they were previously forced to leave a lot more head room before the throttling / shut down point.

The digital thermal sensors have changed all that. No matter what type of software you're running and no matter what part of the core you're stressing, the DTS compares the temperatures of all of the known hot spots on the chip and its output is not the average temperature but the temperature for the hottest point. If any area of the cpu is out of spec then the entire processor is throttled back in MHz by first dropping the multiplier to 6X and it will also try to reduce the core voltage if the motherboard is willing. If the temperature of the processor can still not be controlled then sooner or later it will simply shut down.

All my testing can be summed up quite simply. People worry way too much about the temperature of their processors when there is absolutely no reason to. If you're not overclocked then it will likely take a CPU fan failure before you'll ever get up to the throttling or shut down point.

If you're overclocking then you are much more likely to lose stability before you ever reach the throttling point. Damaging a core processor by overheating it is very unlikely. Intel has put 110% into the thermal management of these processors and they have been specifically designed to shut down before any heat related damage can occur.

The other big concern is long term damage or life of the processor. The boxed Intel processors have a 3 year warranty. Three years from now processors will definitely be cheaper and you'll most likely be richer so $100 for a new one shouldn't be a major concern. I still have a well overclocked Pentium III running strong many years later. It also lived through a stupid user pulling the heatsink and fan off because he was curious :rolleyes: with no long term negative effects.
Click to expand...

I know how the thermal diodes and how the thermal management of these chips work. I've seen it in action too. I've seen fan failures on Pentium 4's and thermal throttling in action. I've never even seen a Pentium 4 or newer processor ever reach shutdown thresholds. What I don't see is anything specifically stating that it's ok to run one of these chips at temperatures above 65c when the Thermal Design White Papers say it's not. I understand what you mean by the mobile processor being able to handle temperatures up to 100C and I get the rest of it. What I am hearing though is speculation on your part that just because the mobile processor are rated to do it, then the desktop parts must obviously be as well. While there is some solid reasoning here, I can't help but play devils advocate because we do not know what criteria Intel uses to seperate mobile parts from desktop parts in advance. Obviously wattage is a consideration, but maybe some cores are considerably more tolerant of heat than others? Maybe those cores are binned as mobile parts and since they are harder to produce, that's the reason for the increased price of the mobile parts. I can speculate too, but the fact remains that the thermal design white paper is the only official document here stating what temperatures a thermal solution needs to keep a processor at to be within specifications.

No document or link provided has told me beyond a shadow of a doubt that the Core 2 Quad can withstand temperatures in excess of 65c for long periods. When the current documentation that I've seen says 65c is the maximum these processors should be kept at and all my efforts on air cooling and even with overclocking have produced results that give me temperatures no where near 65c. I can't help but conclude that keeping things under 65c under load is a good idea.
 
Yes, I would think the lower the temps the better ;)

uncle webb, that's some impressive temperatures you reach, you crazy mofo :D I'd be pretty scared just to turn my computer on w/o a heatsink.

I have an X2... think I'd be capable of doing the same thing? hah
 
Hey, anyone worried about temps can keep running C2D's cooler than they apparently need to and sleep easy with the piece of mind that gives them.

Me, I've pretty much entirely stopped worrying about temperature as a factor in overclocking, it's simply a non-issue for anyone running pretty much any decent after market cooler. Because other limiting factors besides temperature prevent me from running stable LONG before temperature comes up as the throttle.
 
Dan_D said:
What I don't see is anything specifically stating that it's ok to run one of these chips at temperatures above 65c when the Thermal Design White Papers say it's not.
Click to expand...
You're still trying to compare a case temperature guideline, 65C, with a core temperature reported by software like CoreTemp. They are not the same.

INTEL: The TDP specification should be used to design the processor thermal solution. The TDP is not the maximum theoretical power the processor can generate.
Click to expand...

If Intel states that the thermal design power (TDP) has nothing to do with the maximum theoretical power the processor can generate then the 65C thermal design guideline has nothing to do with the maximum safe core temperature for a C2D. That is determined solely by the on chip digital thermal sensors (DTS).

INTEL: The DTS is the preferred method of reading the processor die temperature since it can be located much closer to the hottest portions of the die and can thus more accurately track the die temperature and potential activation of processor core clock modulation via the Intel Thermal Monitor.
Click to expand...

INTEL: The system designer is required to use the DTS to guarantee proper operation of the processor within its temperature operating specifications.
Click to expand...

INTEL: As an output, PROCHOT# (Processor Hot) will go active when the processor temperature monitoring sensor detects that the processor has reached its maximum safe operating temperature.
Click to expand...

PROCHOT# is activated when the DTS reports zero. My E6400 is at an absolute core temperature of 85C when DTS=0. Intel clearly says that is the "maximum safe operating temperature."

Dan_D: There's nothing wrong with your personal decision to limit your core temperature to 65C but the Intel documentation you provided isn't for core temperatures and doesn't back up your decision.

I'm in total agreement that running your processor as cool as possible is a good thing but beyond that I guess we'll just have to agree to disagree.

you crazy mofo :D
Click to expand...
I don't disagree with that!

I'd be pretty scared just to turn my computer on w/o a heatsink.
Click to expand...
Why do you think the picture is a little blurry when it got up over 84C? I was scared s-h-i-t-less and shaking like a dog!
 
You must log in or register to reply here.
Back
Top