Did I fry my CPU?

[SweeJ]

I installed World Community Grid/BOINC a month ago, and I noticed that during when the PC was inactive the client would run in the background to use free CPU cycles to do extra work.

I noticed my CPU temps would go up to 70-72C consistently.

Today, after a crash, rebooting to Windows or even to BIOS, my PC would black out after a few mins and return to boot up screen.

Does this mean World Community Grid/BOINC fried my CPU?

My CPU is an i5-2600.

Thanks for your help in advance.

 

[ch33zw1z]

http://ark.intel.com/products/52213

It could be. Do you have any warranty or maintenance on the box?

 

[Torn Mind]

70-72 degrees Celsius is not enough to kill a Sandy Bridge chip.

Sandy Bridge had boot loop issues, although without knowing your mobo, it's a wild guess as to what is the cause.

 

[VirtualLarry]

70-72 degrees Celsius is not enough to kill a Sandy Bridge chip.

Agreed. What Brand / model / wattage power-supply do you have in there? The added load might have pushed your PSU over the edge, if it was a cheapy or an OEM one.

I do distributed computing, on a number of PCs, but I always make sure I have good CPU / case cooling, and decent high-quality power-supplies installed.

The only case of defective hardware that I've seen, was a burned-out CPU fan on an E-350 all-in-one motherboard rig. Those have tiny, weak, CPU fans, that were not designed to run at top speed 24/7 for months at a time. I have since replaced the fan (twice), and I no longer run DC on E-350 rigs.

 

[Bubbleawsome]

I doubt it. I average 78c when folding, and it hasn't damaged it in nearly a year of (spread out) DC work. I'm betting on an issue with the psu, but maybe a BIOS issue. (Hint; Dell put an OEM 350w, 300w real life, psu in a PC that needed a 400w psu. Died within a year of light gaming.)

 

[SweeJ]

70-72 degrees Celsius is not enough to kill a Sandy Bridge chip.

Sandy Bridge had boot loop issues, although without knowing your mobo, it's a wild guess as to what is the cause.

my mobo is Gigabyte™ GA-Z68AP-D3 rev. 2.0.
PSU is Corsair™ HX620 620W Modular SLI
both are less than 2 years old, self-built PC
Windows 7 Home

i had to Google what "boot loop" means. instead of exact booting timings, it would randomly reboot (usually within 5 mins) with no BSODs, be it from 5 mins of using Windows, Windows starting up, idle log-on Windows screen or even in idle in the BIOS.

if a PC was rebooting even from idle in BIOS, what would usually be the cause? i've never encountered anything like this before, let alone other forms of crashes, from the less than 2 years of using my PC.

 

[SweeJ]

The only case of defective hardware that I've seen, was a burned-out CPU fan on an E-350 all-in-one motherboard rig. Those have tiny, weak, CPU fans, that were not designed to run at top speed 24/7 for months at a time. I have since replaced the fan (twice), and I no longer run DC on E-350 rigs.

fan is stock Intel CPU fan, no over clocking done in BIOS ever nor in Windows. the PC is mainly a 24/7 office desktop multitasker with occasional extremely light gaming.

i let my PC run idly in BIOS bcos i wanted to monitor temps, but all 4 cores didn't go past 55C. however, that's just 5 mins of running idly, as the system would boot again.

 

[SweeJ]

I doubt it. I average 78c when folding, and it hasn't damaged it in nearly a year of (spread out) DC work. I'm betting on an issue with the psu, but maybe a BIOS issue. (Hint; Dell put an OEM 350w, 300w real life, psu in a PC that needed a 400w psu. Died within a year of light gaming.)

the odd thing is i've been using this system for less than 2 years, and the heaviest tax on the system is constanly maxed out use of my 8GB RAM, as i run several programs with plenty of browsers in the background. however, CPUs are often idle until last month i installed folding client.

BIOS is current version since i built the PC nearly 2 years ago.

my PSU is Corsair HX620.

so you think folding caused my PSU to be faulty?

 

[Bubbleawsome]

I doubt that too... Have you tried the trusty memtest? Have you tried running who crashed to see if there is any long made?

 

[mfenn]

fan is stock Intel CPU fan, no over clocking done in BIOS ever nor in Windows. the PC is mainly a 24/7 office desktop multitasker with occasional extremely light gaming.

i let my PC run idly in BIOS bcos i wanted to monitor temps, but all 4 cores didn't go past 55C. however, that's just 5 mins of running idly, as the system would boot again.

So by "won't boot", do you meant "won't boot into Windows" or "won't POST"?

 

[Bubbleawsome]

So by "won't boot", do you meant "won't boot into Windows" or "won't POST"?

I think it auto reboots after around 5 minutes in windows. That's what I've gathered.

 

[SweeJ]

I doubt that too... Have you tried the trusty memtest? Have you tried running who crashed to see if there is any long made?

i have tried memtest and the Windows Memory Diagnostic Tool (the option you get after you his F9), unfortunately, the tests aren't able to complete as around the 5-10 min mark, the system will just reboot.

 

[SweeJ]

So by "won't boot", do you meant "won't boot into Windows" or "won't POST"?

my PC would auto-reboot in all conditions.

meaning, after around 5-10 mins, it would just auto-reboot with no error notifications or screens.

i am able to log into Windows, BIOS, Safe Mode, etc. every option I choose to do, seems to run normally. i would even try System Restore, and it would run with no problem, that is until around 5mins.

however, after around 5mins, the system would black out and reboot.

does that sounds like a mobo, CPU, memory or PSU issue?

 

[SweeJ]

I think it auto reboots after around 5 minutes in windows. That's what I've gathered.

it even auto reboots around 5 mins in BIOS (ie. that blue screen set up by hitting Del key when they system starts up), with me not changing anything in BIOS, ie. just letting it be idle.

 

[Bubbleawsome]

Whocrashed is near instant, and it tells you what crashed. Defo worth a shot.

 

[Torn Mind]

It is possible that the PSU is not getting a "Power good" signal after 5 minutes and hence the motherboard "resets". Something is off with the power delivery system, and since this is prevalent, although with some variations in the symptoms, on quite a few boards with that model name, it is not just an isolated issue.

You might want to disable PLL overvoltage, since one guy who wrote a review at the board's Newegg was able to fix the issue, although it is not guaranteed to work.

http://www.newegg.com/Product/Product.aspx?Item=N82E16813128538

 

[SweeJ]

i'll have to find the receipts of the old PSU and mobo, since both those seem to be the main culprit.

 

[mfenn]

I'm betting on PSU issue, but mobo and memory are also possible. The memory is easy enough to test by running with a single stick at a time (try different slots).

 

[Torn Mind]

I'm betting on PSU issue, but mobo and memory are also possible. The memory is easy enough to test by running with a single stick at a time (try different slots).

I would pin it on the mobo, as the boot issues are not an isolated issue on that specific model board.

 

[SweeJ]

just thought i'd update this post.... so i've been really busy for the past few months to really sit down to trouble shoot and find replacements.

after replacing the motherboard, the boot looping was still there. this time, i noticed the cold boot loop were getting sooner and sooner after each subsequent boot. until..... it just wouldn't boot up, and there was a strong burnt smell.

so off to the store again to get a new PSU; no more Corsair for me, wouldn't have thought distributing computing software tipped it over. and i don't even game much.

after installing the Seasonic PSU, no more cold boots!

thanks to all again, for your help.

 

[Bubbleawsome]

Seasonic makes the corsair 620w.

 

[SweeJ]

Seasonic makes the corsair 620w.

oh yeah.... that's right.... they do..... darn. lol!

 

[Bubbleawsome]

They are widely considered the best PSU OEM, so it's no issue. But yeah, pretty ironic.

 

[Puffnstuff]

If that ps is still under warranty you should rma it.

 

[SweeJ]

yeah, i will check the warranty, and if it's still alright, i'll rma it.