[SOLVED] 8700k died after professional delidding

[soulfetsh]

I wanted go get my CPU delidded after temps started to peak at 99C while gaming, stayed above 80C. I blamed it on the AIO liquid cooler first, got that replaced. But temps stayed high above 90C.

I got the delidding done professionally. After I received my CPU back, temps dropped to 61C under full load PRIME 95. I watched it for 5 minutes thinking that all my problems were finally solved. I left the computer with the PRIME95 running for 4 hours. When I came back the monitor was showing black screen. I forced a shutdown by pressing the power button. When I started the computer again it never posted. CPU LED on the motherboard turned up, and error code showed 00.

I bought a new 9700k immediately from Microcenter. It works. I bought a cheap motherboard to test the 8700k, it didn't work. 9700k also worked on the new motherboard.

9700k idles at 29C at stock settings. If I run the PRIME 95 it reaches 99C just like my old CPU. Stock settings were trash (pushing the voltage to 1.45v). So I overclocked to 5Ghz at 1.35v. Temps are around 70 while gaming.


Cooler: H115i RGB pro XT
Ram: 32GB 3000Mhz CL15
Motherboard: ASUS Maximus Hero X WIFI
GPU:6800 XT

I just wanted to share my experience. I am not sure if I could have prevented the CPU failure if I was there to watch the PRIME95 test.

 

[punkncat]

It would seem to me that simply seeing those temps to begin with that something was already 'in process' for the CPU failure.

 

[SamirD]

I think your cooler is the root cause--shouldn't be hitting 99C to begin with.

 

[Reclusive Eagle]

"Professional" means nothing.
As a photographer I've seen "Professional" "world renowned" photographers in South Africa deliver the most garbage work I have ever seen complete with water marks and a zebra border for a paying client... Turns out "world renowned" means he went to America once and was featured in a budget magazine.

Go get your money back

 

[soulfetsh]

I think your cooler is the root cause--shouldn't be hitting 99C to begin with.

If I disable AVX insructuins temps start around 85C. Slowly increase and peak at 91. But with AVX instant 99C and thermal throttle.

 

[soulfetsh]

It would seem to me that simply seeing those temps to begin with that something was already 'in process' for the CPU failure.

Is this how a CPU typically dies? Am I this unlucky that making the decision to delid and CPU dying happened at the same time?

 

[SamirD]

If I disable AVX insructuins temps start around 85C. Slowly increase and peak at 91. But with AVX instant 99C and thermal throttle.

Too hot. Check cooler.

 

[SamirD]

Is this how a CPU typically dies? Am I this unlucky that making the decision to delid and CPU dying happened at the same time?

My guess is that it already took a beating running under that excessively hot condition and the delid helped, but it's still stressed so it finally died.

 

[Mr.Spock]

"professional" delidding could describe Geek Squad work so take it with a grain of salt...
might've been a faulty chip, but it voided the warranty so we'll never know.

certain versions of Prime are not recommended for specific Intel processors overstressing the FPUs, usually they just throttle though - however your setup may have allowed for catastrophic failure before shutdown...

 

[soulfetsh]

Could a "faulty chip " work fine for 3.5 years and than fail?

 

[Mr.Spock]

chips could fail over time, and due to heat or extreme voltage. delidding may have just exacerbated the process or it could've been damaged in the process.

 

[TerryLaze]

Stock settings were trash (pushing the voltage to 1.45v)

Your main issue is this here, reported 1.45V could be 1.5 or even more in reality, but even 1.45 is enough to cause degradation after enough time.

High Vcore together with lifted power limits and a fan curve that is adjusted for quiet and not highest cooling = huge chance of something breaking.