Sign in with
Sign up | Sign in
Your question

I7 2600k failure

Last response: in Overclocking
Share
April 29, 2012 10:13:08 AM

Had my rig running 4.5 ghz fine running below the 1.4v limit, decided to push it, had it run 8 threads at a bit below 1.7v, reverted to normal, ran fine for ages.
recently decided to push to the 1.5v limit and clocked 4.9ghz 4 threads.
woke up one morning with it blue screening on me everytime. downclocked it to standard and got a boot and login shortly before blue screen again.
So i decided to downclock it, had it running longest at 2.5ghz 4 threads but it would inevitably fail, blue screened quicker at lower clocks
I changed my approach to the downclock and limited the number of cores used, by running 3 cores w/o HT i was able to easily reach a clock of 4.0ghz without any voltage change without any issues of blue screening.
I think I may have blown a core if that is even possible?
Temperature has never been a problem and has only ever reached 80 under extreme clock stress tests.
I have of course tried a reinstall of windows with and without my common drivers/programs etc but the same problem exists.

Just wondered if anyone had an idea as to what may have happened?
Thanks in advance.

Current (working) setup:
i7 2600k 4GHz 3 cores WC - Gigabyte Z68X-UD5-B3 - HD 6950 - Asus Xonar Essence STX - 8GB G.Skill Ripjaws 2133MHz - OCZ Agility 3 120GB SSD - WD Caviar Black 1TB

More about : 2600k failure

a b à CPUs
a b K Overclocking
April 29, 2012 11:03:17 AM


:o  How high you over volted your CPU's IMC (memory controller). A very high IMC voltage can kill a CPU's IMC instantly. There is also a possibility of the CPU burned by trying 1.7V and 80C temp is definitely not low at such a very high overclock. At 4.0+ ghz speeds temp of the CPU should be 70C or lower and also check your RAM on other machine to ensure its working properly. Since you are using an SSD drive there is also a possibility that it failed and you may try installing the OS on to a HDD.
April 29, 2012 11:11:57 AM

spp85 said:
try installing the OS on to a HDD.



Have already installed os onto another hard drive, same problem. Im fairly sure ive blown one of the cores somehow as it runs fine on just 3, but didn't really know if that was possible.

Just to note; it only ran into the 80degrees on a 5.1GHz OC with Hyper Threading under prime95 stress testing, at which point i thought i might underclock it a bit. and stick with my original goal of 4.5GHz.
Related resources
a c 138 à CPUs
a b K Overclocking
April 29, 2012 12:12:06 PM

TheBrutalFlan said:
Have already installed os onto another hard drive, same problem. Im fairly sure ive blown one of the cores somehow as it runs fine on just 3, but didn't really know if that was possible.

The reason why all overclocking guides warn that there is a chance of frying your CPU is because many people have indeed ended up with dead CPUs from pushing their luck too far, accelerated aging or cooling failure.

With built-in thermal protections added to CPUs over the past ~10 years the risk is much lower but still exists.
April 29, 2012 12:40:48 PM

InvalidError said:
The reason why all overclocking guides warn that there is a chance of frying your CPU is because many people have indeed ended up with dead CPUs from pushing their luck too far


Yh, i knew the risks, buy the ticket, take the ride. Just thought it was weird that it works fine when limited to 3 cores. I always thought if you blew a processor thats it, no more processor.
a c 138 à CPUs
a b K Overclocking
April 29, 2012 2:26:28 PM

TheBrutalFlan said:
I always thought if you blew a processor thats it, no more processor.

Depends on what fails and how. You can have a single-bit failure in an execution unit that is used only for a few uncommon instructions that allows the computer to run normally until software happens to execute one of these instructions on the affected core, you can have a transistor failure that shorts internal supply rails and causes surrounding material to melt until the VRM/PSU either fails or trip their over-current protection, you can have a single transistor failure on a critical common signal path that brings the whole CPU to a halt (memory and IO interfaces for example) or anything in-between.

If one non-critical failure caused the whole CPU to fail, Intel wouldn't have been able to sell Pentiums with bad cache as Celerons and AMD would not be able to sell Phenom X4 with a defective or under-performing core as X3s. Sometimes the disabled resources can be successfuly hacked re-enabled because they were disabled to meet SKU demand on an otherwise perfectly fine chip, sometimes the hacks fail because the resources were genuinely defective.

Failures are not necessarily fatal if the affected circuitry and failure mode are non-critical, can be disabled or bypassed.
April 29, 2012 3:42:27 PM

InvalidError said:
Depends on what fails and how. You can have a single-bit failure in an execution unit that is used only for a few uncommon instructions that allows the computer to run normally until software happens to execute one of these instructions on the affected core, you can have a transistor failure that shorts internal supply rails and causes surrounding material to melt until the VRM/PSU either fails or trip their over-current protection, you can have a single transistor failure on a critical common signal path that brings the whole CPU to a halt (memory and IO interfaces for example) or anything in-between.

If one non-critical failure caused the whole CPU to fail, Intel wouldn't have been able to sell Pentiums with bad cache as Celerons and AMD would not be able to sell Phenom X4 with a defective or under-performing core as X3s. Sometimes the disabled resources can be successfuly hacked re-enabled because they were disabled to meet SKU demand on an otherwise perfectly fine chip, sometimes the hacks fail because the resources were genuinely defective.

Failures are not necessarily fatal if the affected circuitry and failure mode are non-critical, can be disabled or bypassed.


well it seems to be running fine on 3 cores so hopefully its a single core issue. No more 5.0GHz clocks for me though. Just wanted a bit more info on what may be happening, cheers for that.
April 30, 2012 10:10:11 AM

http://click.intel.com/tuningplan/ <-- look into this it might help you and get you a replacement 2600k for 25$, one time no questions asked replacement offered by Intel for the K-series.

Also, 1.7IMC overcooked, damn you are lucky to not have a complete blowout.
a c 103 à CPUs
a c 239 K Overclocking
April 30, 2012 12:18:36 PM

TheBrutalFlan said:
Had my rig running 4.5 ghz fine running below the 1.4v limit, decided to push it, had it run 8 threads at a bit below 1.7v, reverted to normal, ran fine for ages recently decided to push to the 1.5v limit and clocked 4.9ghz 4 threads.

Temperature has never been a problem and has only ever reached 80 under extreme clock stress tests.


Just wondered if anyone had an idea as to what may have happened?


1.7v and you wonder what happened?

Temperature is irrelevant, too much voltage can fry your CPU instantly, you fried part of it.

It is crippled now, it will never run right again.



a c 186 à CPUs
a c 150 K Overclocking
April 30, 2012 3:24:27 PM

Once you go above 1.52v for an excess amount of time. Your entire system blew up!
!