Question CPU died suddenly, what are the possible causes ?

[LakeShowJr]

I have a rig with a Ryzen 5 2600X that has been running a little over 4 years. It was cooled by a DeepCool 212 cooler and has never been OC'd, and I have never seen it get hot (36°C idle - 50's when I used to game). Today I was watching a youtube video and suddenly got a black screen, followed by a reboot. The PC got stuck in boot loop. I tried to boot off a secondary SSD that has Windows 11 on it, thinking my primary SSD was toast, but even with the secondary SSD as boot drive, the PC was still going through the reboot loop.
I figured it's the CPU or the MB.

I had an old Ryzen 1300X laying around, popped it in, and sure enough the PC booted into Windows.

My question is, what can cause a CPU to just die suddenly? Again, it was never OC'd and stayed cool. One thing I will mention is that I always have a lot of tabs open in Firefox and Chrome. After a week or so, the PC would bog down (become slow) as the memory usage would be at 90-95%.
However, the CPU would still be at low usage. All I had to was restart Firefox/Chrome, and the PC would operate normally again.

So, is it just bad luck? or should I be looking at something else?

The PC is connected through a CyberPower UPS and has not experienced power interruptions.

 

[ex_bubblehead]

Unlikely to be a CPU failure as it attempts to boot.

 

[CountMike]

I have a rig with a Ryzen 5 2600x that has been running a little over 4 years. It was cooled by a Deep Cool 212 cooler and has never been OC'd, and I have never seen it get hot (36°C idle - 50's when I used to game). Today I was watching a youtube video and suddenly got a black screen, followed by a reboot. The PC got stuck in boot loop. I tried to boot off a secondary ssd that has windows 11, thinking my primary ssd was toast, but even with the secondary ssd as boot drive, the PC was still going through the reboot loop. Figured it's the CPU or the MB. Had an old 1300x laying around, popped it in, and sure enough the PC booted into windows.

My question is, what can cause a CPU to just die suddenly? Again, it was never OC'd and stayed cool. One thing I will mention is that I always have a lot of tabs open on firefox and chrome. After a week or so, the PC would bug down (become slow) as the memory use would be at 90-95%; however, the CPU would still be at low use. I would restart firefox/chrome, and the PC would operate normally again.

So, is it just bad luck? or should I be looking at something else?

The PC is connected through a CyberPower UPS and has not experienced power interruptions.

Just like any electronic device, CPU can die sudden or gradual death but by your info it looks more like Windows issues or with memory. Heat and OC are also unlikely cause for failure, it take many years of extreme abuse for a modern CPU to deteriorate or die.
Try changing CMOS battery and update BIOS first.

 

[Viking2121]

Yeah something is not stable or of the fritz, I would place my bet on the ram, because another CPU works with it doesn't mean its 100% stable, it might just work with the 1300x because it forced the bios to reconfigure for the 1300x.

Cmos, can reset it, if you go to replace the cr2023 battery, might as well just reset the cmos/bios. Unplug the system from the wall, hold the power button down to drain power, pull the battery, and let the system sit for a good 10min to be sure, replace the battery or install the old one if you didn't buy a new one with the positive side up, usually has writing on the positive side. See if it boots with the 2600x.

If it doesn't, try setting your ram speed down to 2400mhz and see what it does, if it works, increase the ram speed 1 step at a time, Ram may have degraded a bit from age, that can happen.

Good Luck!

 

[Amdlova]

Overclocked Dram? my 5700G died with 6 months with a low speed 4133mhz ddr4 =) fried a 2400T memory after complete dies

 

[LakeShowJr]

I tried the original cpu with a single stick of ram of a different brand, it still exhibited the reboot cycle.

don't think it's windows as I tried booting off two different SSD's, and the problem presisted.

I just find it weird the way the CPU died. I remember watching a music video on youtube, I switched to another tab to check email, as soon as I clicked back on the video tab, bam! black screen!

Anyways, I have another CPU coming today from amazon, will update.

 

[Eximo]

I would attempt to reflash the BIOS as a first step, rather than just clear the CMOS, could just be simple corruption which allows the older CPU to boot and the newer one to no longer be properly started.

From some of the channels I have followed. Ryzen 2000 and 3000 chips have a habit of the I/O die simply dying or losing a memory channel suddenly.

Could be plenty of causes. Interposer connection failure from many heat cycles. Or too tight of a CPU cooler install.
Unregulated temperatures on a part of the die with no temp sensors (Something AMD has made note of several times)

 

[logainofhades]

Only time I had a CPU die was when a crap power supply died an took the CPU with it.

 

[geofelt]

Processor failures are rare, but it does happen.
If you work well with a different processor the original could have failed.

Separate the boot to windows from the boot to bios.
Run memtest86+
It boots from a usb stick and does not use windows.
You can download it here:

Memtest86+ | The Open-Source Memory Testing Tool

Memtest86+ is an advanced, free, open-source, stand-alone memory tester for 32- and 64-bit computers (UEFI & BIOS supported)

memtest.org

If you can run a full pass with NO errors, your ram should be ok.
Your basic hardware should be ok.

If you can run memtest, then your issue is windows related.
With the working chip, update the motherboard bios to currency.

 

[LakeShowJr]

Cleared CMOS by popping the battery and putting a screwdriver to the two cmsclr terminals for 15 seconds. Cmos cleared, but no change, PC still stuck in boot loop.

By the way, when I say boot loop, I mean it reboots after POST, it does this twice, and then the screen goes black. I have to manually shut it off at that point and turn it on again to repeat the process.

I made some changes to the CPU voltage. It was running on auto at 1.3v. I changed it to 1.1v, and upon trying to boot, it got to a blue screen with Error 0x0000428. After about 10-15 seconds the PC turned off.

 

[LakeShowJr]

Processor failures are rare, but it does happen.
If you work well with a different processor the original could have failed.

Separate the boot to windows from the boot to bios.
Run memtest86+
It boots from a usb stick and does not use windows.
You can download it here:

Memtest86+ | The Open-Source Memory Testing Tool

Memtest86+ is an advanced, free, open-source, stand-alone memory tester for 32- and 64-bit computers (UEFI & BIOS supported)

memtest.org

If you can run a full pass with NO errors, your ram should be ok.
Your basic hardware should be ok.

If you can run memtest, then your issue is windows related.
With the working chip, update the motherboard bios to currency.

I don't think it's windows. as I have explained in previous posts, I tried to boot off a different SSD drive with windows 11 installed, and PC was still displaying the same issue.

I have also tried the original cpu with a single stick of different memory.

 

[LakeShowJr]

Just for the hell of it, I tried to boot into Win10 installation media that I created on a flash drive, and the PC reboots upon trying to boot off the flash drive.

 

[BFG-9000]

Electromigration will eventually kill some critical part of every processor even if other parts might last hundreds of years--and as Intel found with 13th and 14th gen Core, pushing nearly the same voltages as the 45nm chips from 15 years ago used through the latest Intel 7 process node can accelerate this process.

I ran across a dead 600nm Pentium 1 chip with bad L1 cache. It still worked just fine with the L1 cache disabled in the BIOS, but that made it slower than a 386. Without the ability to disable parts of the processor one at a time like that, you can't hope to figure out what part broke much less why. At most you'd be able to discover the iGP died or something, except your chip doesn't have one. You could try your single stick in the other channel to see if a whole RAM channel in the memory controller died.

It's important to note MTBF is not expected average lifespan, no matter how large the sample size.
If you have 1 million 20-year old humans and 1,250 of them died over the course of a year (probably mostly from not looking where they're going while texting), the failure rate is 1,250/1,000,000 or 0.125% a year.
MTBF is the inverse of failure rate so thus calculates out to 1/0.00125 or 800 years.
But the life expectancy (remaining service life) of those people cannot be expected to come anywhere near the MTBF

 

[LakeShowJr]

UPDATE:
Got the new 5600x in and PC booted up into windows. I went ahead and did bios update to the latest version and chipset/drivers update while I'm at it. System has been running for about an hour without problems.

The 5600x is running a little hotter (2600x would run at 38c @Idle and the 5600x runs at 42-45c @Idle) I get occasional fan ramping up now and then. I used Corsair thermal paste, hopefully it'll run a bit cooler once the paste spreads across the cpu.

Now I'm off to start the RMA process on the 2600x

 

[beyondlogic]

I have a rig with a Ryzen 5 2600X that has been running a little over 4 years. It was cooled by a DeepCool 212 cooler and has never been OC'd, and I have never seen it get hot (36°C idle - 50's when I used to game). Today I was watching a youtube video and suddenly got a black screen, followed by a reboot. The PC got stuck in boot loop. I tried to boot off a secondary SSD that has Windows 11 on it, thinking my primary SSD was toast, but even with the secondary SSD as boot drive, the PC was still going through the reboot loop.
I figured it's the CPU or the MB.

I had an old Ryzen 1300X laying around, popped it in, and sure enough the PC booted into Windows.

My question is, what can cause a CPU to just die suddenly? Again, it was never OC'd and stayed cool. One thing I will mention is that I always have a lot of tabs open in Firefox and Chrome. After a week or so, the PC would bog down (become slow) as the memory usage would be at 90-95%.
However, the CPU would still be at low usage. All I had to was restart Firefox/Chrome, and the PC would operate normally again.

So, is it just bad luck? or should I be looking at something else?

The PC is connected through a CyberPower UPS and has not experienced power interruptions.

what is the brand of motherboard just because it boots with ryzen 1300x doesnt mean jack 1300x uses less power and is a quad core.

thats to not say its not a cpu but its insanely rare.

most of the time its the motherboard more to due to power delivery it could also be the power supply not delivering enough power on start up.

 

[beyondlogic]

UPDATE:
Got the new 5600x in and PC booted up into windows. I went ahead and did bios update to the latest version and chipset/drivers update while I'm at it. System has been running for about an hour without problems.

The 5600x is running a little hotter (2600x would run at 38c @Idle and the 5600x runs at 42-45c @Idle) I get occasional fan ramping up now and then. I used Corsair thermal paste, hopefully it'll run a bit cooler once the paste spreads across the cpu.

Now I'm off to start the RMA process on the 2600x

what paste are you using ? also not uncommon for 5600x to run a bit warmer

 

[LakeShowJr]

paste is Corsair TM30

 

[LakeShowJr]

what is the brand of motherboard just because it boots with ryzen 1300x doesnt mean jack 1300x uses less power and is a quad core.

thats to not say its not a cpu but its insanely rare.

most of the time its the motherboard more to due to power delivery it could also be the power supply not delivering enough power on start up.

The MB is an ASRock B450M Steel Legend

 

[beyondlogic]

paste is Corsair TM30

i personally use ARCTIC MX-6 granted to get it to spread i use a vinyl glove non powdered using just my finger to navigate the paste all over the die till die isn't visible then i put the cooler on doing this method achieves a even surface over the die and is better then the pea method which for ryzen isnt recomended.
cleaning before doing the above.
to get the cpu clean i keep it in the stocket locked in and use isopropyl alchohol 99 strenght and a cotton buds. i do this to the cooler till the cpu and cooler give only clean reactions on the cotton buds then i know its gotten all of the micro stuff.
ryzen 5700x is my cpu and im sitting at operating 4 tabs open at 32c max ive ever gotten it to go above 65c.
i assume your using coolermaster 212 which is a good cooler but the nuts at the rear of the motherboard can destroy a motherboard if tightened ( i did this once with a gigabyte board)

Thermalright Burst Assassin 120 CPU cooler would be a good upgrade. as it gives 2 extra heatpipes.​

 

[beyondlogic]

The MB is an ASRock B450M Steel Legend

the vrm or power delivery on that board gets quite hot and needs decent airflow.

 

[Viking2121]

Only time I had a CPU die was when a crap power supply died an took the CPU with it.

Man the only time I've ever had a CPU die was my 5820k, 3 of them in fact and a 12 core Xeon, Was on the Asus x99 Deluxe, Asus took the board back twice and nothing was wrong in the testing they did, then they was pretty rude to me when I contacted them for the 3rd time. The first 5820k was OC to 4.5ghz, nothing to crazy, the other 2 and the Xeon all stock. I find out later Asus OC socket was in fact killing Haswell and Broadwell-e CPU's and Asus even screwed most of them people over and brushed it under the rug.

I moved to a 5960x and a Evga x99 Classified with the same PSU, Ram, storage, in fact that system is still going to this day. I love that EVGA board.

Out of 20 years of working on PC's that was the only system I've ever had where the CPU's were actually dying or dead, I've never really seen another dead CPU since what 2015? Seen CPU's with messed up memory controllers but worked non the less.

 

[Misgar]

My question is, what can cause a CPU to just die suddenly?

One possible answer is ESD (Electro Static Discharge). CPUs are susceptible to static electricity damage due to mishandling.

When I worked for billion dollar international corporations, they spent millions equipping all their labs and assembly areas with "anti-static" flooring, work benches, tools, etc. When handling components, boards and fully assembled customer equipment, we were obliged to wear ESD shoes, ESD lab coats and ESD wrist straps or gloves.

You don't want all the flight navigation systems to stop working when you're flying at 33,000ft, 10,000m, because some idiot zapped all the chips with static during assembly. In a home computer, it doesn't matter.

When you built your PC four years ago, did you place the motherboard on an ESD mat and use an ESD wrist strap, or did you simply touch the metal case and trust to luck before handling the CPU? It'll never happen to me!

Was your CPU brand new in the manufacturer's sealed box, or was it second hand from eBay "protected" in ordinary bubble wrap (horrible nasty static generating c**p)?

ESD damage to CPUs, RAM and motherboards is insidious. A quick static zap of 100V+ and your prized component might (and I stress might) die before you've retired the computer.
https://www.esda.org/esd-overview/esd-fundamentals/part-1-an-introduction-to-esd/

Unless you have an electron microscope, you probably never know the truth. CPUs are remarkably resilient, but once in a blue moon they drop dead for no obvious reason.

https://blog.item24.com/en/esd-en/identifying-esd-damage-using-an-electron-microscope/

https://automotivetechinfo.com/2016/12/preventing-electrostatic-discharge-damage/

 

[nvfan99]

The fact that you can get into the BIOS but it crashes trying to load Windows, which is quite CPU intensive, might indicate your 2600X has degraded - much like what's happening with the Intel 13th/14th gen CPUs lately, it just took a lot longer for it to happen to your CPU. The fact that you indicated it crashed even faster after undervolting also makes this seem more likely.

Solutions to survive loading into Windows with the chip may be to either lock the CPU boost frequencies to a lower MHz number, or increase voltages to the CPU until it's stable, as likely the Vmin necessary for stability has gone up over time.