**Official e8400/e8500 Retail OC Thread

[necron66]

nice explanation there

check out what i also found in that CPUz dump
(btw 0x34 was the corresponding value as per your explanation which is also mentioned in my screen under Core0)

http://img82.imageshack.us/img82/598...board01xw8.png

Mine reads same as core temp

Hardware monitor
-----------------------------------------------------

Intel Core 2 Duo E8400 hardware monitor

Temperature sensor 0	42°C (107°F) [0x3F] (core #0)
Temperature sensor 1	46°C (114°F) [0x3B] (core #1)

Dump hardware monitor

but it never moves from here at idle, only when I load it up.

[Fpat]

Nice, what cooling do you use?

SS

[ArCElM]

Did you measure voltage by voltometer, or in BIOS?
I need set in BIOS 1.485V , MCH1.45V, VTT 1.31V.
My Voffset is 0.085V + Vdrop 0.032V- so to Run at 4GHz I need 1.368V
After 1.485Vcore and restarting comp - PC Health Status showing me 1.39Vcore - I didn't mesure voltometer.

Sorry I do not have a voltmeter, but After 1.465Vcore and restarting comp - PC Health Status shows me 1.34Vcore - nasty - I am 100% stable at 4.1Ghz - with borderline temps at 63C ORTHOS cpu stress test.

[Merc14]

I shot off an email to The Inquirer last night regarding heat problesm with the e8400's I stated that the sensor diodes aren't up to snuff. They posted this today. http://www.theinquirer.net/gb/inquirer/new...-reported-intel

There have probably been 100 emails o them but I said I'd spare them more details as they should do a search on "e8400 temperature problems" for an eye opener. Seems that is exactly what they did.

[unclewebb]

There have been a few reports of sensor diodes getting stuck at idle and not going any lower but all of them go higher once you apply full load. Intel never designed these sensors to report idle temperatures. They designed and individually calibrate each one to trigger thermal throttling as the DTS approaches a value of zero and there isn't one report yet of that not happening.

The only problem with these sensors is that a programmer or two didn't read the Intel documentation and is using the DTS for a purpose that they were not designed for.

I've yet to see one single report of a new E8x00 processor over heating and melting because the DTS was faulty and thermal throttling and shut down wasn't triggered. We need to put this problem in the proper context.

[Merc14]

There have been a few reports of sensor diodes getting stuck at idle and not going any lower but all of them go higher once you apply full load. Intel never designed these sensors to report idle temperatures. They designed and individually calibrate each one to trigger thermal throttling as the DTS approaches a value of zero and there isn't one report yet of that not happening.

The only problem with these sensors is that a programmer or two didn't read the Intel documentation and is using the DTS for a purpose that they were not designed for.

I've yet to see one single report of a new E8x00 processor over heating and melting because the DTS was faulty and thermal throttling and shut down wasn't triggered. We need to put this problem in the proper context.

I think folks here are more concerned about what the actual temperature of the cores are at any given time so they can determine if they are hurting the chip long term rather than is it getting ready to melt. 99.9% of the people who buy a PC with an e8400 in it have no idea that they can read its temperature. Nor do they care. But for those that push these chips to extremes, it is important that they be able to judge the heat load on the cores at any given setting so as to determine where they want to settle thier everyday overclock.

I think you are exactly correct that Intel will not RMA these chips because they are functioning perfectly as far as the spes are concerned. As you said, no one's chip has melted down because of a bad diode and that is the diode's only real function. Intel will take note, however, of widespread anger among the enthusiast class and hopefully address our desire to have accurate temperture readngs available at all times. Of course, with AMD fading into the sunset, there will be little or no incentive for Intel to do anything as they will be the only game in town for high performance CPUs.

[Blackwhite]

Sorry I do not have a voltmeter, but After 1.465Vcore and restarting comp - PC Health Status shows me 1.34Vcore - nasty - I am 100% stable at 4.1Ghz - with borderline temps at 63C ORTHOS cpu stress test.

Your voffset is bigger then main.

But now I've got another problem - I can't logg to the windows, it is black screen - yesterday was fine - any ideas?

[ArCElM]

No idea sorry.

But one thing I would like to know, from anybody, what then, is the detramental effect of the voffset - should I be considering I have my BIOS set to 1.465? or should I be paying attention to what the chip is actually getting as informed by PC Health Status - which should concern me.

Thanks.

[unclewebb]

The enthusiast community may be rather vocal but they make up a very small percentage of Intel's sales.

The reason Intel went to a DTS system for processor temperature control was to get around the problems with bios based temperature sensors. Problems and returns of the first generation Core 2 Duo desktop processors must have been small enough that they feel comfortable and have raised the TjMax value with the E8x00 series.

Everyone needs to let go of absolute core temperature values. It's not an important number. Maximizing the value returned by the DTS will maximize the amount of temperature headroom you have before thermal throttling and will also increase full load stability. If everyone would report the DTS when comparing their results like Intel intended then there wouldn't be a problem.

Here's a good paper from Intel about the DTS:
http://irevues.inist.fr/bitstream/2042/6579/1/TMI23.pdf

The DTS is calibrated at manufacturing conditions and the reference point is set to this test temperature. Functionality, electrical specifications and reliability commitments are guaranteed at maximum Tj as measured by the DTS. Any test inaccuracy or parameters variance are already accounted for in the DTS set point.

[cadaveca]

I did an experiment...my e8400 never budges off of 40c. Not load, not idle...cpu reported temp is 35c(not DTS). Again, this temp never changes.

Anyway, took my pc outside...-30c outside.

Chipset is watercooled also...

Chipset dropped down to -5c.

Cpu temp and DTS never moved from 35c and 40c, respectively.

So, i go back inside, let the system warm to room temp, ensure there is no condensation...

remove cooling, put on stock cooler....ah....


WTF!!!! DTS 75c loaded on stock speed and volts? CPU 70c?

BTW, my cpu has a "t" at then end of FPO.

[ArCElM]

Everyone needs to let go of absolute core temperature values. It's not an important number. Maximizing the value returned by the DTS will maximize the amount of temperature headroom you have before thermal throttling and will also increase full load stability.

MATE I may be being thick but wtf does this mean?

I mean if you say your on 65C LOAD OR (105 - 65) 40DTS LOAD, it IS the exact same thing right? .. im rly ocnfused on this one, what AT ALL is the difference here, and yes I am well aware I am clueless, so help me, plz thanks.

[PytonOrm]

holy crap my e8400 does this fsb on air with asus commando and bios ver 1605

http://valid.x86-secret.com/show_oc.php?id=306427

606Fsb and stil testing.... Ownage

[unclewebb]

WTF!!!! DTS 75c loaded on stock speed and volts? CPU 70c?

Is it possible that you didn't use any thermal paste when putting on the stock cooler or didn't push in very many of the cheap push pins? These processors will get red hot, even at default MHz and voltage, when there is a problem with how the heatsink is attached.

It definitely sounds like the DTS sensors on your chip are having the "getting stuck" problem. Why not complain to Intel and get them to agree to give you a new chip if your present chip ever dies due to overheating. That way you'll be able to run the hell out of your chip with no worries.

With my E6400 I found that when overclocking to the max (~3600MHz) that I couldn't get the DTS below 25 for very long without it randomly rebooting. If I left 25 degrees of headroom then I was Orthos stable for hours at this speed.

This sounds crazy but overclocking my chip forced me to keep it cool and I never have to worry about it over heating. The processor will simply shut down long before it ever reaches or even gets near the thermal throttling or shut down stage. I'm sure Intel is well aware of this safety feature, whether they intended to do this or not.

[vera]

holy crap my e8400 does this fsb on air with asus commando and bios ver 1605

http://valid.x86-secret.com/show_oc.php?id=306427

606Fsb and stil testing.... Ownage

Any p5k-e users ? reach 600fsb with this 45nm cpu's ?

[unclewebb]

I mean if you say your on 65C LOAD OR (105 - 65) 40DTS LOAD, it IS the exact same thing right?

Sorry for confusing the issue. You are absolutely right, it is the same thing. The only difference is everyone has preconceived notions about what a safe absolute core temperature is to run there processor at. A core temp of 60C scares some people and for other people the magic number is 70C or 80C or whatever before they say, "That's just too frikkin hot."

What Intel says is that as long as the value of the DTS is a positive number then you are operating it safely and within their intended design specifications. When overclocking, for maximum stability, we all need to keep as far away from DTS=0 as possible. Everyone running around comparing absolute temperature values that are not even documented by Intel is pointless, especially when Intel fully guarantees their processors at TjMax.

Functionality, electrical specifications and reliability commitments are guaranteed at maximum Tj as measured by the DTS.

No one here is even close to that point so as long as you're stable, absolute temperatures just don't matter anymore.

[synna]

holy crap my e8400 does this fsb on air with asus commando and bios ver 1605

http://valid.x86-secret.com/show_oc.php?id=306427

606Fsb and stil testing.... Ownage

Any more info on your voltage settings? I have the same setup.

[ArCElM]

Sorry for confusing the issue. You are absolutely right, it is the same thing. The only difference is everyone has preconceived notions about what a safe absolute core temperature is to run there processor at. A core temp of 60C scares some people and for other people the magic number is 70C or 80C or whatever before they say, "That's just too frikkin hot."

What Intel says is that as long as the value of the DTS is a positive number then you are operating it safely and within their intended design specifications. When overclocking, for maximum stability, we all need to keep as far away from DTS=0 as possible. Everyone running around comparing absolute temperature values that are not even documented by Intel is pointless, especially when Intel fully guarantees their processors at TjMax.

Functionality, electrical specifications and reliability commitments are guaranteed at maximum Tj as measured by the DTS.

No one here is even close to that point so as long as you're stable, absolute temperatures just don't matter anymore.

THANKS! I fully understand now, but mate! thats a BOOOLD STATEMENT! when my CPU hits mid-late 60's I must admit, I think - RIGHT! thats as far as I am going without better cooling!

Remember longevity is an issue also, there has to be SOME weight to the mentality that, your CPU is HAPPIER and healthier when kept at a nice 60-70C rather than running it at 75C+ daily..

[Zeus]

These Wolfies are rather strange, i can run 3d mark and even SuperPi 32M at 4.5GHz with 1.53V but it hits a brick wall around 4620MHz, no matter the voltage i throw at it.

Still 32M at 4.5GHz is nice.



Untweaked, so maybe sub 11 minutes is possible.

[Xvys]

Here is Packo's email to Intel posted in the OCForums E8400 overheating thread:
-------------------------------------------------------------------

I e-mailed Intel abut the sensor problems
Heres what intel said about senors problems, talk about a load.

First,
My e-mail, by the way I did copy and paste some info from this thread to get things started.

the temp probs in the E8400 are stuffed. The Coolest (the guy who
makes coretemp) reports that according to Intels spec, his program is
reading the temps correctly. In other words, coretemp is reading what
the cpu is telling it. Case and point, my E8400. According to
coretemp, one core is locked at 43c, the other at 51c. They will not
read any lower than this, ever... Once the temp moves above these
values its starts to rise, but other report completely frozen probs
(sometimes just reading 7c!). My motherboard reports 34c on the cpu
all the time, i have never seen it move off this value (this may be
fixed in future bios though). Nevertheless, there seems to be
problems with the sensors being reported on many websites, can you
confirm this and when will the recall take place?

-------------------------------------------------------------------
Intels response,

The maximum operating temperature of the Intel(R) Core(TM) 2 Duo
processor E8400 is 61.4 degrees Celsius. As long as the processor is
operating under this temperature it is operating within
specifications. We recommend setting any thermal alarms about three
(3) degrees below the maximum recommended temperature for the
processor. We do not have a normal operating temperature for the
processor as this temperature will vary depending on the chassis and
other hardware installed on the system as well as the actual load the
software is placing on the processor.

You do not need to worry about the temperature of each core, you only
need to pay attention to the CPU temperature itself.

Currently we do not have any report on sensor problems with the
Intel(R) Core(TM) 2 Duo processor E8400.

[emoners]

so Intel is saying that CoreTemp didn't upgrade it's version to measure the new Wolves???

[unclewebb]

Remember longevity is an issue also, there has to be SOME weight to the mentality that, your CPU is HAPPIER and healthier when kept at a nice 60-70C rather than running it at 75C+ daily..

You would think so but Intel doesn't seem to agree. They offer a full 3 year warranty on their chips and have designed them so they can run reliably up to the point where the DTS = 0.

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.

Intel's definition of the PROCHOT# signal clearly shows that you can run your processor right up to TjMax and you are still operating it safely. If it wasn't safe to operate a desktop C2D at this high a temperature then Intel would lower TjMax and not let users run their processors so hot.

The processor contains Digital Thermal Sensors (DTS) distributed throughout die. These sensors are implemented as analog-to-digital converters calibrated at the factory for reasonable accuracy to provide a digital representation of relative processor temperature.

"Reasonable accuracy" means that CoreTemp and other programs shouldn't be using this data to report an exact absolute temperature because the sensors themselves are not exact.

They provide reasonable accuracy of relative processor temperatures. That means if DTS=40 at full load and then you slap on a different cooler and DTS=50 then your processor is running approximately 10C cooler. If DTS=40 during the winter and then becomes DTS=30 in the summer then you'll know that your CPU is running about 10C hotter during the summer. It provides relative data to the factory set throttling point for your individual processor and shouldn't be used when comparing absolute temperatures from one processor to the next. Intel doesn't even document that TjMax is a fixed value across the entire E8x00 line so why is temperature monitoring software assuming that it is?

[4Qman]

My max super pi 1mb is around 5.4 and 32mb was 5.3 but i was rushing and have dead ram so this may have caused errors at these speeds.

On Cascade

[unclewebb]

It says right on the Intel website the Thermal Spec for the E8400 and it's not 61.4C. It would be nice if they could hire people that could actually read their own spec sheets.

Thermal Specification: 72.4°C
http://processorfinder.intel.com/det...px?sSpec=SLAPL

Also keep in mind that this is a case temperature reading and is intended for heatsink designers. It is measured at the geometric center on top of the IHS and has absolutely nothing to do with the sensors that CoreTemp is reading.

[Mickey Padge]

You do not need to worry about the temperature of each core, you only
need to pay attention to the CPU temperature itself.

So the read out of CPU temp by a program like Abits uGuru, should be what I look at and be concerned over, not each individual core, via core temp software? At least that is the opinion as far as Intel are concerned?

[cadaveca]

Is it possible that you didn't use any thermal paste when putting on the stock cooler or didn't push in very many of the cheap push pins? These processors will get red hot, even at default MHz and voltage, when there is a problem with how the heatsink is attached.

Nah, no chance of that, unfortuantely. First tiem removing the cooler from the box, and the imprint was as good as it gets. I'm just blown away that the cpu has not throttled at all. I guess I must be pretty close to that point tho, on stock cooling.

Maybe that's why the "T" in my FPO...who knows.