Is it rare to get a DOA CPU?

[Sniper82]

Is it rare? I've never had it happen. Also how likely is it for a CPU to die on down the road as long as you don't use extreme voltages on it and keep it temps safe?

Reason I ask is I am thinking about buying a new CPU off something sealed for $45 less than retail and am wondering if its worth the risk. Anyone know if Intel requires proof of purchase/Invoice copy to RMA?

 

[Duvie]

I have never gotten one....I have probably bought 20 retail chips, 10 oem chips, and about 4 used chips on the forums...never had one bad...Only ever had to rma 2 cpus, 1 of which I later found to be the motherboard but I had no real way to test then other to follow the beep codes...

 

[Sniper82]

Originally posted by: Duvie
I have never gotten one....I have probably bought 20 retail chips, 10 oem chips, and about 4 used chips on the forums...never had one bad...Only ever had to rma 2 cpus, 1 of which I later found to be the motherboard but I had no real way to test then other to follow the beep codes...

I see you have a few Gigabytes. Curious is the DS3 worth the extra $20 for the solid caps ? I am looking at getting one of the two.

 

[Duvie]

Originally posted by: Sniper82

Originally posted by: Duvie
I have never gotten one....I have probably bought 20 retail chips, 10 oem chips, and about 4 used chips on the forums...never had one bad...Only ever had to rma 2 cpus, 1 of which I later found to be the motherboard but I had no real way to test then other to follow the beep codes...

I see you have a few Gigabytes. Curious is the DS3 worth the extra $20 for the solid caps ? I am looking at getting one of the two.

So far from the chips I have...I say no....My E6600 actually oc'd better in the S3...unknown why...

Until I test my QX6700 I cannot say for sure...

The one thing I can say is it is my belief all S3's have the newest chipset revision, but not all DS3's do...Now anandtech said they saw no diff in ocing, but i hear rumblings that this may affect quad core support...Again I will be testing that shortly

 

[bjp999]

I have never heard of a chip arriving DOA, although have heard of a few Conroes that won't OC hardly at all. Very rare posts and still not convinced it's not user error.

I had a Northwood P4 3.4 go bad and RMA'ed it. It was about a year ago, and the CPU was almost 3 years old. I had the receipt but they didn't ask for it. I assume that the ID numbers from the CPU told them it was still under warranty, although it was somewhat close. The process was painless, Intel was very accomodating and shipped very fast. They did not ask many questions about how I knew it was the CPU and not something else. I think I filled out an online form but may have spoken to them also - don't really remember.

My CPU appeared to work but I started getting random BSODs and reboots. I bought a new MB thinking that had gone bad. Problems got worse. Local shop offered to test it (they said they had never seen a broken CPU and I don't think they believed that the CPU was really bad), but in the end they agreed. I RMA'ed the CPU (and returned the MB) and the new CPU has worked perfectly in my old MB. Right or wrong I blame the problem to the stock HSF. I bought a Zalman cooler to replace it.

The warranty was great because otherwise I'd have had to find a socket 478 processor on eBay or start over with a new socket T MB. I'm not sure if Intel would sell you a replacement for a product that was out of warranty.

 

[genec57]

Over the past 15 years I have built hundreds of PCs. During that time I have had one Intel die after about three months and I have had a couple of DOA AMDs (these were older chips. Zero in the past 10 years.

 

[NuAlphaMan]

In the past 10 -11 yrs that I've been building systems, I've yet to get a DOA CPU.

 

[The-Noid]

Never DOA. I blew a Northwood up though. Intel gave me a new one even though I told the guy I Was overclocking it. They seem to be pretty accomadating on the high end stuff. I believe if you told them you overclocked an X6800 they would probably still give you a new one.

 

[Sniper82]

but no one had been asked to include copy of invoice/receipt when RMAing? I know their warranty terms says original purchaser.

 

[sonoran]

Originally posted by: Sniper82
Is it rare? I've never had it happen. Also how likely is it for a CPU to die on down the road as long as you don't use extreme voltages on it and keep it temps safe?

If it's in an unopened box your chances of getting a DOA are probably below 1000:1. The chips are tested pretty thoroughly before packaging and shipping.

As for overclocking, I don't think it'll make any significant difference as long as you stick to stock voltage.

 

[Sniper82]

thx alot I think I will go ahead and get it since it is $45 below retail. I can put that toward a better board or something else.

But no one had been asked to include copy of invoice/receipt when RMAing? I know their warranty terms says original purchaser.

 

[Damn Dirty Ape]

I've worked in a pc repair shop and installed hundreds (200-350) maybe and never had one doa. This includes both AMD and Intel. Unless there is damage in shipping or user error/mistake.. Never had to RMA any that failed on their own, always something else taking it out (mobo, PSU).

 

[lobbyone]

Its pretty hard to shake loose all those pins when shipping anyway, unless they ship the chip in just an envelope with no protection, on topic, never had a DOA chip as well YESSS!

 

[Sniper82]

how many has had CPU die down the road o/cing or not and was it due to, to much voltage?

 

[bjp999]

Mine was completely stock. No OCing. (I've since learned the dark arts. )

 

[quadcore]

Never had a DOA CPU but I did receive a DOA RAM stick once.

 

[Ronnie]

Never had a DOA, only chip I killed was poor insulation job on phase. My poor little opteron running at 3.5ghz

 

[TankGuys]

It's *very* rare. I've sold thousands of processors, and only had a handful returned supposedly DOA (though, in truth, I'd guess they were just bad OCers )

 

[Keitero]

Seen it happen. Had a customer bring back a server because the machine would lock up when trying to load an SMP kernel of RHLE.

 

[Markfw]

Only chip I ever had die was a Xp1700 that had the power supply go out, and take the cpu and motherboard with it !

 

[lopri]

This thread makes me curious. We know we don't get a DOA CPU, and know they're long-lasting. Why can't other parts (in a system) have such quality/durability?

 

[Zenoth]

Originally posted by: lopri
This thread makes me curious. We know we don't get a DOA CPU, and know they're long-lasting. Why can't other parts (in a system) have such quality/durability?

I still use a P4 2.4C which is OC'ed to 2.8, with DRAM and a Radeon 9500 Pro. All those still work 100% stable and do the job (just for basic text work and some old games for my father). If you keep good maintenance on your hardware, not just CPU's, then it will usually also last quite long, if not virtually for a life time (well maybe that'd be a little too far fetched, but perhaps 8, 9 or 10 years wouldn't be exaggerated).

The oldest CPU I owned was a VIA C3 700 Mhz, and even though it was extremely slow for gaming it worked properly. I eventually sold it for $5 to a friend, and last time I heard about it (maybe 4 or 5 months ago) the system with it worked well.

 

[BadThad]

Originally posted by: bjp999
I have never heard of a chip arriving DOA, although have heard of a few Conroes that won't OC hardly at all. Very rare posts and still not convinced it's not user error.

I had a Northwood P4 3.4 go bad and RMA'ed it. It was about a year ago, and the CPU was almost 3 years old. I had the receipt but they didn't ask for it. I assume that the ID numbers from the CPU told them it was still under warranty, although it was somewhat close. The process was painless, Intel was very accomodating and shipped very fast. They did not ask many questions about how I knew it was the CPU and not something else. I think I filled out an online form but may have spoken to them also - don't really remember.

My CPU appeared to work but I started getting random BSODs and reboots. I bought a new MB thinking that had gone bad. Problems got worse. Local shop offered to test it (they said they had never seen a broken CPU and I don't think they believed that the CPU was really bad), but in the end they agreed. I RMA'ed the CPU (and returned the MB) and the new CPU has worked perfectly in my old MB. Right or wrong I blame the problem to the stock HSF. I bought a Zalman cooler to replace it.

The warranty was great because otherwise I'd have had to find a socket 478 processor on eBay or start over with a new socket T MB. I'm not sure if Intel would sell you a replacement for a product that was out of warranty.

WTH? I build/repair systems and I have a customer that had a bad 2.4C. In order to get an RMA from Intel, I had to test the system with a different processor to prove the 2.4C was dead! Also, I had to give Intel the cpu description AND serial number of the "proof" cpu. The whole Intel RMA process was THE MOST PAINFUL OF MY COMPUTING EXPERIENCE! Not to mention it took them like 3 months to get the replacement to me.

On a side note, that 2.4C system had EVERY component replaced before the CPU. It would give random BSOD's. I was completely shocked it came down to the CPU....but the new CPU permanently fixed the system from BSOD's!

 

[quadcore]

Originally posted by: lopri
This thread makes me curious. We know we don't get a DOA CPU, and know they're long-lasting. Why can't other parts (in a system) have such quality/durability?

Hint: There are no electrolytic capacitors in microprocessors. No moving parts either such as on a hard drive. Unless you count the fan.