than why make unlocked CPU at all, really strange to say that but the i5 2500 can only go up 4x and don't go above 80 C. using prime 95
That is a good point, but there may be some confusion here. The original post was about a non-K, locked CPU. The response by the Intel person may be a typo on "unlocked". Then again, what is the warranty limit of an unlocked CPU, endless multipliers of 50+? The apparent Catch 22 is, yes it is unlocked, but OC at your own risk. Which does make sense. What manufacture of any "hot rod" part can offer an open ended warranty? But how do you prove that a part was operated beyond it's limits? A CPU should shut down if it over heats, but will it do the same when to much voltage is applied? Honest OC enthusiasts know they can and do kill CPUs, and don't expect them to be replaced under warranty.
I can set the Turbo multipliers on my i7-2600k to 45, and it will run for long periods under testing at ~4.5GHz, but I leave all voltage setting in the EFI on Auto, and don't use over voltage settings. I also keep EIST and C1E enabled, and have good after market air cooling (which actually voids the warranty, I believe.)
We really need to find the definitive policy statements regarding these situations, I don't have them available right now, time to scan the fine print.
i have some real problem i guess... my system is over heating alot and i dont know why, i built a system with core i5 2500... the specs are
core i5 2500
4GB ADATA 1333 RAM
The casing air flow is good in the sense that that casing has no side cover also the tests i ran were taken at about 30C room temperature...
the thing is that when i run prime 95 the CPU temperature goes very high very quickly... just after starting the program the temperature jumps from 40C idle to about 60C and withing 2 minutes the temperature reaches about 88C the PC does not shutdown though and if the program runs continuously the temp keeps on rising... i didnt let it run that long for fear of overheating the system... I am using stock cooler...
can you guys tell me if something is seriously wrong or not???
i was rendering a 3DSMAX project the other day and the temps remained at 85C for the entire time... it took almost 30 minutes to render the scene...
also in compilation of programs the samething happens... however, the PC never shutdown itself for overheat protection though...
can you guys tell me whats wrong??
other thermal readings during the tests are
Mine's a lot like yours Hassaan. I'm running at default in the bios, no overclocking, using stock heatsink. My temps run high 40s idle to 83c under 100% load (Prime95). That's pretty damn hot I'd think! My room temp is 20c. I contacted Intel about it and was told this was "normal", and under normal use its rated for 72c, peak of 100c (in reality, mine does 98 according to CoreTemp). For example, the processor in another computer (AMD) runs 32c/53c. So, it runs at load what this one runs idle! I've bought a Coolermaster Hyper 212+ to see if I can bring those temperatures down. I'll report back.
Great! As you have now seen, the standard Intel CPU cooler is simply not effective enough for long term use under very high load situations. While it has become standard procedure for many enthusiasts to run CPU load testing programs like Prime95, that type of "usage" is not at all typical of the average user, which is the usage model that Intel uses to specify their coolers capabilities. That is, long periods of low usage punctuated by short to medium duration instances of moderate to high loads, moderate meaning not more than a five minutes, IMO.
I also believe that enthusiasts become carried away with running CPU load tests at or approaching 100% usage literally for hours. Imagine doing the following to your car: Start the car in your garage and leave it in park. Take a brick and put it on the accelerator pedal so it is pressed to the floor board, also known as wide open throttle. Then, walk away and go to bed and let it run like that over night. Return in the morning, and if the car is still running, think "Ok, good, my car is fine". But if the car over-heated, was damaged, or shut off (BSOD), think "oh great, my car is not stable, something is wrong with that engine...". Is "testing" of this kind really reasonable?
My apologies for this speech, and this is not aimed at anyone in this thread, but is simply my opinion.