Cpus dying more often now?

[Justintime]

i5 8500 - Was in some random HP Deskpro, BSODs - Critical process died, IRQ less than blah blah etc.. windows won't install without bsods when it did manage to complete would lock up etc... tried memory, gpu etc.. nothing. Tried a new 8500 and all problems gone. put the old one back, bsods. Tried in a different board with more control and noted with turbo boost off cpu was 100% stable, enabled and it went back to bsod country.

9400F - Same thing really except disabling multi core solved it so probably a bad core.

9900 non K - Won't run dual channel memory in any board with any memory, usb transfers in the 30-50mb/sec range but with good cpu in the 100mb + range with same stick and files.

9700K - Very random POST and lock up imminent but worked fine otherwise with no BSOD etc.. lol, cleaned it to hell but same result on multiple boards.

i5 7500 - Died randomly in clients pc, new cpu works fine. Never seen that amount of cpus die in a decade muchless a year! Seen a few Ryzen ones died too but those are mostly DOA, are Cpus much more fragile these days or what? Most of these cpus are non overclocked stock ones too.

 

[Tetras]

Wow. That hasn't been my experience, found the 8th/9th gen as reliable as any other. From what I can recall, the 11th gen had an abnormally high reported failure rate.

 

[Justintime]

Oh yeah work with few clients who have hundreds of Coffee, Kabylake etc.. with no issues but i hardly look to the cpu being the fault at first these just surprised me being they came from such mundane scenarios too lol

 

[Tetras]

Oh yeah work with few clients who have hundreds of Coffee, Kabylake etc.. with no issues but i hardly look to the cpu being the fault at first these just surprised me being they came from such mundane scenarios too lol

Yeah would be more understandable if they were massively overclocked/overvolted, but non-ks in oem builds is weird.

 

[Tee Hee Johnson]

Nope.

 

[arknor]

I had a 4670k for about 8-9 years and never really had any issues.
although towards the end I would get 1-2 cpu errors in event viewer every week, but never had any noticeable issues.

I put the vcore up a notch and they went away again, ran at 4ghz the whole of it's life

 

[Glanza]

Nope, only time I've seen a CPU fail in the last 10 years is when people have installed it and either bent the pins on the chip (AMD) or damaged their LGA socket on Intel. Which is less of a chip failing and more user error breaking it.

 

[Cyber-Mav]

most likely mobo socket warping or dodgey pins in board? maybe reinstalling cpu would have had same effect as new cpu?

 

[Justintime]

most likely mobo socket warping or dodgey pins in board? maybe reinstalling cpu would have had same effect as new cpu?

Different boards, same results.. new cpu sorted it every time. I even compared to capacitors/smds on the back etc.. to see if anything was missing.

 

[Robert896r1]

Tortured my 9900k since day 1 and still runs at the same exact voltage as ever. Not even 5mv of deg.

 

[Jumper118]

Sounds like an Intel problem. I suspect with all the cpus running at 95-100c now the current gen ones won't last too long either.

 

[Cyber-Mav]

Hmm hot running cpus with lower lifespan, seems plausible since now cpu warranties are down to 2 years.

 

[Tetras]

Hmm hot running cpus with lower lifespan, seems plausible since now cpu warranties are down to 2 years.

Huh, is this an AMD thing? Intel 13th gen docs say three year, from what I can see?

 

[humbug]

Touch wood in 25 years of owning PC's i have never had a CPU or GPU die on me.

My mother got my old Athlon X2 5200+, that thing was used every day, it was 15 years old when she replaced the PC, it was still working just fine at the time, painfully slow for the modern internet, but still in good health.

 

[humbug]

Huh, is this an AMD thing? Intel 13th gen docs say three year, from what I can see?

Some Nvidia GPU's now have a 2 year warranty, i just checked Zen 4, they are 3 years.

 

[Rroff]

Aside from a small number of first generation Ryzen CPUs which seemed to have some kind of production defect I've not seen any change or increase in CPU problems with more recent CPUs. Pretty much all CPU problems I've encountered have been down to the end user to be honest - a common one spilling a drink into the machine then for some reason not admitting it and claiming it just suddenly started working funny.

 

[eddiew]

I can only speak of my 1 sample, but my 8700k lost it's ability to hold 5ghz at any sane voltage a few months back. Now running 4.5 stable. For now.

(It's a lot cooler and lower voltage though, and I'm not sure I can tell the difference.)

 

[ltron]

I had an I7 940 that would randomly reboot and lose a memory channel before eventually recovering, but I'm not a hundred percent certain whether it was the CPU or the motherboard although I strongly suspect the former; it was a real pain to troubleshoot. All my other CPUs have been fine including my current AMD R9 3900X. It's the motherboard that tens to fail or be faulty in my experience.

 

[Rroff]

It's the motherboard that tens to fail or be faulty in my experience.

Can be a pain at times - I've found some models of motherboard where even tiny amounts of flex in the board can cause random errors - some older Asus boards where the southbridge chip would very easily lift slightly off some of its BGA points with the slightest flex for instance due to poor solder and others where a slightly heavier than normal GPU could do it, etc. often very hard to narrow down what the fault is.

 

[13oots2]

The only CPU I have had die was down to operator error. Had been up all night sorting out my Duron 900, a friend came around and distracted me. Put the system back together and fired it up, could hear a random ticking sound, it was the processor slowly blowing itself to bits as I had forgotten to mount the heat sink, at least it meant I could now get an Athlon 1gHz.