Max Temperature for an i5 2500K

jon_k

Gawd
Joined
Jul 29, 2004
Messages
567
Running an i5 2500K with an Asus P8Z68V-Pro/Gen 3 motherboard cooled with a Thermalright Ultra 120 heatsink attached to a 120mm fan.

Right now I have the CPU at 4GHZ @ 1.24V. Prime95 stress testing hovers around 48C and Intel Burn test hovers around 55C.

I played around last night and I ran it at 4.5GHZ @ 1.3V. Prime95 temps were mid 60s and Intel Burn Test was high 70s with a few quick peaks in the low 80s. I played some games at this voltage/speed and actual usage temps were still really low (high 40s, it didn't even crack 50C). 100% stability in all cases.

My question is what are the max temps I ought to run this thing at? I have a feeling I could push it to 4.7 - 4.8 but I don't want to roast my chip. Thanks.
 
I think all the Sandys start throttling at 95C. Most people try to keep temps below 80C I think though.
 
Should I aim for below 80C running Prime 95/IBT, or below 80C with actual usage?
 
I try to keep my 2600k below 80C under FULL load, but thats personal preference. If your keeping your chip at 80C under normal usage, under full load it would probably be well above 95C. At the end of the day when your overclocking most Sandys, heat usually isn't a problem. Aim for the lowest voltage you can to get the job done, weather its a huge overclock or just a small one.
 
My question is what are the max temps I ought to run this thing at?

~100C is the maximum. The chip will automatically throttle if you get a little over this. If you get a lot over that it will shut off completely. As for safety I would not go much over 90C for daily usage.
 
80C or lower with IBT (which heats my 2500k much more than Prime95). My experience w/the 2500k has been that I hit voltage limitations before temps get crazy.
 
I read that Intel specs a lot of these chips at 73 or 71 or around there. Is 80 a normal value?
 
I must have gotten a really good chip. I reduced the voltage a bit, and @ 1.24V I'm running my i5 at 4.7GHZ. Temp is 66C after Prime95 stress testing. Solid as a rock. Kind of scared to do Intel Burn Test lol. I'm going to see if I can push it a little higher on that voltage.

Another quick question. Why is my VCore different from the CPU voltage I set? For example i set the CPU voltage to 1.24 but my VCore is 1.424. I have PLL-Overvoltage enabled (was told this helps with OCing) so maybe that has something to do with it.
 
Your actual voltage to the chip can be read in Windows from CPU-Z or any Hardware monitor program. Your actual chip voltage is 1.424v (VCORE), not 1.24. When you enable PLL-Overvoltage, and LLC in the bios, it allows the chip to decide how much more voltage it needs. [email protected] would be a very good O/C on little voltage, [email protected] is about average. Most people w/25-2600k's need 1.4v+ to get to 4.7-5.0Ghz. My chip will do 5Ghz easily @1.46v, but that's more voltage then I want to run 24/7. I settled on [email protected] with mine. Temps run 66-68C under full load.
 
The max Tcase according to Intel is 72c, so you'll be able to run at around 80c and maintain that.

If you hit 90 or 100c for a short time that's fine, but chips start to degrade really quickly at those temps.
 
Tcase has nothing to do with the operating temp of the processer. You can read about Tcase here
But yes, I would definatly keep my chip at 80C or lower. Temps that high usually are not a problem with the SB's though. You'll usually run out of voltage long before heat is an issue with the Sandys.
 
Tcase has nothing to do with the operating temp of the processer. You can read about Tcase here
But yes, I would definatly keep my chip at 80C or lower. Temps that high usually are not a problem with the SB's though. You'll usually run out of voltage long before heat is an issue with the Sandys.

Uhh..., what?

Tcase absolutely has to do with the operating temperature of the processor. Intel specifies what the max Tcase is, in this case 72c.

You can usually go 10c or so hotter than this, so if your core temps are 80c you are probably nearing the max Tcase.
 

Tcase
: Technically this refers to the temperature that you can measure using a thermocouple embedded in the centre of the heat spreader, but this is only done in the factory by Intel or by users willing to drill their heat spreaders open to insert a thermocouple. Therefore Intel provided a diode between and below the cores with a reading calibrated by the BIOS which can be used instead. This reading can vary greatly when the BIOS version is changed, but will not necessarily change if the BIOS calibrations were not altered between versions.

Software supplied by motherboard manufacturers typically reads only this sensor, not the DTS. It is often labelled “CPU.” SpeedFan usually labels this as “CPU” as well.

Tcase Max and Thermal Specification:
Can be found on the Intel Processor Spec Finder. This is often confused by users as the maximum temperature for the cores, but in actual fact it refers to the maximum temperature that should be reached by the Tcase sensor. This specification is rounded to a relatively precise 0.1 C.

Perhaps read the link I gave before posting?
 
I think you are completely misunderstanding what I'm saying. If not, read the part you highlighted in red.

The maximum temperature the tcase temperature should hit, in this case 72c, happens when the cores get 10c or so hotter then it. This is where the 80c max temp comes from. At 80c your reaching the max temperature the tcase sensor should hit.

Edit:

To make what I'm saying clearer. The tcase sensor is usually 10c or so cooler than the core temp
 
I managed to push the chip up to 4.8GHZ at 1.25V (actual VCore was 1.424). Did an IBT and it hit 90C. Was stable, but no way I'm running the chip that hot. Turned it to 4.5GHZ and temps were under 80C for all my benchmarks. Changed LoadLine Calibration to medium so it only overvolts to 1.384.

Much cooler :)
 
I think you are completely misunderstanding what I'm saying. If not, read the part you highlighted in red.

The maximum temperature the tcase temperature should hit, in this case 72c, happens when the cores get 10c or so hotter then it. This is where the 80c max temp comes from. At 80c your reaching the max temperature the tcase sensor should hit.

Edit:

To make what I'm saying clearer. The tcase sensor is usually 10c or so cooler than the core temp

Tjunction is what matters. That is the temperature when the CPU begins to throttle, and for SB chips that is 98C. The temperatures that Coretemp, Realtemp, etc, read are in relation to Tjunction. Forget about Tcase completely - it is impossible to measure, and is only used by heatsink manufacturers to develop their products. According to the specs, you could run the CPU at 97C (as measured by CoreTemp) 24/7 without any ill effects.

The relationship between Tcase and Tjunction is dynamic and depends on the cooling solution, so you can't just say it is 10C cooler. 80C is just a rule of thumb people use for a decent temperature for stress testing, it doesn't have anything to do with the Tcase.
 
Last edited:
Back
Top