Computer Repair Techs - How often have you seen bad/failed processors?

[Pickles von Brine]

I work as a repair tech (been one for a few years now) in the consumer market. I have ran into 2 failed processors in about a year or so. First one was with my own personal system. My Ryzen 3600X just died. Black screen, fans kick up, no post. Doesn't help the X570 Aorus Elite has no post LEDs on it and I didn't have a speaker... Second processor to fail was with a client (just fixed). He had a Ryzen 3600. Would fail under low loads. Bios update did not fix this issue. At first thought it was a power supply, didn't work. Then replaced the board. Didn't work. Then finally tried an old Ryzen 1400 and system worked fine. Replaced with a new Ryzen 3600. No issues. 

How often have you guys seen processors fail? Honestly, I cannot think of too many times. Maybe 5-6 years ago, i would see some AMD processors fail in laptops, but even then it was unusual. 

[Semper]

If my enthusiast level "tech" experience over the last ~15-20 years or so is of any value to you, that number comes to a total of one. if we're including systems that were so heavily outdated that it really didn't matter, that number is ~3-4. I've had some really old hardware die on me out of age, but it was expected, not running anything important.

The only CPU I've had die on me unexpectedly was an FX-8150. It lasted about a year and bit the dust, swapped with an 8350 that's still working last known.

[Levent]

I personally cracked dies of two 3570ks. Other than my own fuck up, never and I worked at a local repair center when I was in highschool. We would get laptops with chipsets or GPUs dead every other day.

[porina]

I don't think I've ever had a CPU I can prove go bad. I did have a laptop fail, and only managed to narrow it down to the main board (soldered CPU) but can't diagnose further.

 

I recall in the P4 era I was worried about SNDS if anyone remembers that. Sudden Northwood Death Syndrome. Was apparently happening often enough to get a name, but I didn't experience it myself, even on my sample I used with a phase cooler.

 

I did have stability problems with a 3700X, but I'm 99% sure that's a mobo thing. It seemed ok one one, but unstable on another. On that board, a positive voltage offset seemed to help. The mobo seemed fine with a 2600 in it. It might just be worst case tolerances meaning that particular sample combo didn't work.

[WhitetailAni]

I have had one...

that has bent pins that aren't possible to bend back without godlike skills.

I broke it.

[jaslion]

I have a small box of em so gonna list the ones I put in a word doc as I don't know where that box is:

 

Pentium 4 2.66ghz x2

Pentium 4 HT 3.06ghz

Pentium 4 2.4 ghz

Pentium 4 530 x3

Pentium D 915

Pentium dual core E2140 x5 (cuz you could oc that poor little 1.6ghz thingy to 3ghz+ no issue but of course people would go overboard with the voltage and poof)

Athlon II x2 250

Amd phenom 8750 x2 (probably tri core to quad core unlock failures)

Core 2 duo e8400

I3 540

Sempron 145 x4 (mining boom got a lot of motherboards from that and these were a very popular choice)

i7 2700k

i5 3550p

Fx 9590

Fx 8320 x2

Xeon e5 2697v2 ( a dud from ebay)

i7 5820k

i5 4440

pentium g3258 x2

Amd 7800k

i5 6500t x11 (office giant aquarium fail systems fried and I got them for almost nothing lots of dead ones obviously)

Ryzen 1400

I9 9980x

Ryzen 2700x

 

[DrMacintosh]

I've never seen a bad chip before. But the amount of PCs/Laptops that I've built/serviced is still pretty small. I would think the most likely thing to see would be bent pins/pads rather than actually dead chips. 

[Slottr]

We just upgraded all of our systems company wide, out of like 4-500 systems, and the ~10 that didnt work, none of them were the cpu

[RAM555789]

I've seen a PSU kill a CPU but never the actually chip fail for a reason other than bent pins by the end user.

[divito]

In my experience as a tech and in personal life, I've had zero "fail."

Now, there have been some that had odd quirks in regards to under-performance, or strangeness in regards to their voltages / clocks that led to certain behaviors. But to outright stop working without something else intervening (like a PSU issue or an overclock)? Hasn't happened to me or in my travels; would just be rare on its own.

[freeagent]

I’ve never killed a cpu with or without voltage. I have only seen one dead cpu.. a dead 3770K that came with my z77 oc formula. It still has conformal pads in the socket area so I can only assume it died in sub zero conditions. Killed a few boards, a few gpus, a few psus but never a cpu lol. I did try once or twice but obviously not hard enough because I’m yella.

[minibois]

Not counting user error (bent pins, pads lifting up, etc.) or obvious damage from other components (i.e. electrical surge), it's quite rare, but it does happen.

It's one of the last things one would suspect to be damaged and I'd say most users will never encounter such damage.

[unsorted]

Where is the 'none' choice?

I haven't seen a single CPU die (pun!).

[nbrowser]

I'm not a "tech" by any means but I do the odd builds for others...that said in the oh, 25 years I've probably built about 200ish systems, only had one CPU go bad on me and that was in one of my own rigs and that was a i7 7700K. Found it odd but yeah, only CPU I've ever seen fail.

[Pickles von Brine]

52 minutes ago, unsorted said:

Where is the 'none' choice?

I haven't seen a single CPU die (pun!).

Added.

9 hours ago, jaslion said:

I have a small box of em so gonna list the ones I put in a word doc as I don't know where that box is:

 

Pentium 4 2.66ghz x2

Pentium 4 HT 3.06ghz

Pentium 4 2.4 ghz

Pentium 4 530 x3

Pentium D 915

Pentium dual core E2140 x5 (cuz you could oc that poor little 1.6ghz thingy to 3ghz+ no issue but of course people would go overboard with the voltage and poof)

Athlon II x2 250

Amd phenom 8750 x2 (probably tri core to quad core unlock failures)

Core 2 duo e8400

I3 540

Sempron 145 x4 (mining boom got a lot of motherboards from that and these were a very popular choice)

i7 2700k

i5 3550p

Fx 9590

Fx 8320 x2

Xeon e5 2697v2 ( a dud from ebay)

i7 5820k

i5 4440

pentium g3258 x2

Amd 7800k

i5 6500t x11 (office giant aquarium fail systems fried and I got them for almost nothing lots of dead ones obviously)

Ryzen 1400

I9 9980x

Ryzen 2700x

 

How? OC?

[Purgent]

Killed two 5820k CPUs.

 

Both were covered under RMA from Intel.

[Guest]

11 hours ago, jaslion said:

I have a small box of em so gonna list the ones I put in a word doc as I don't know where that box is:

 

list snip

 

Yeah, looks like my list  

 

Killed a couple from de-lidding

A couple from lost pins.

Dropped one smashed the corner in.

have ruined 9590 memory controller channel from HT overclocking 4ghz+

Have received a few dead chips in the mail. 

Had a 3.2ghz opteron 165 die as I used it. After 5 reboots clocking even blow stock freq. poof. no more posting. 

Seen a few in this forum for sure. Mostly AMD chips too. 

Think I've killed or damaged a chip in just about every AMD platform lol.

 

Boards too. A good stack of them. Recenty just pitched an M4A79 deluxe (AM2+) it will only post if you clear the cmos.... every single restart. I've tried everything short of baking it for a reflow, but meh. The only fun chips where unlockers and Windsor and Brisbane Athlons imo. AM3 was a different story. 

 

Here's a 4100 I got too hot during a de-lid. It popped! 

 

[Mark Kaine]

been doing this for about half a year. 

 

broken processors : zero 

broken PSUs : a plenty 

broken mobos: more than you'd think! 

broken HDDs : a many 

broken laptop hinge: one... 

[Helpful Tech Witch]

Not a repair tech, but I guess my info might be helpful.
Ive killed a couple laptops cpus, cracked one, gouged the other (dachshunds+ laptops = mess)

Was trying to delid a c2d e8300, took a chunck out of the die with my knife.

dropped a pentium 4 (pga one) and the carpet caught like 1/2 the pins.

and to top it off, I threw a p4 that was fried by psu (like sparks and burnt off pads) into a fire.

[jaslion]

12 hours ago, Pickles - One of the Jar said:

Added.

How? OC?

I repaired systems for years. I don't remember much for most but some were obvious where the psu or board has a catastrophic failure and thus the whole system dies or certain parts.

[TetraSky]

Like, twice. A LONG time ago in the Pentium 3/Pentium 4 era.

And even there, it was usually broken CPU pins.These days it's mostly bent pins in the socket.

 

I do remember once that someone had tried overclocking their CPU too much and basically fried everything by pushing way more voltage in there than necessary. 

[NinJake]

Over the last 5 years, not a single CPU has failed on a computer.

[r.perez]

Not a repair tech but I've been an enthusiast for 22 years. I have never seen a CPU that failed on it's own.