i7-6900k fail

Discussion in 'Intel Processors' started by stormslayer, Dec 19, 2016.

  1. stormslayer

    stormslayer [H]Lite

    Messages:
    69
    Joined:
    Apr 26, 2005
    I'm freaked out. Have built almost two dozen systems over the past 15 years and I've had the usual amount of hardware fails -- a few motherboards that had issues, a failed gpu, etc. Never had a CPU go bad.

    Built a new rampage V ed. 10 w/ i7-6900k about three weeks ago and this week in the middle of reading my email (very demanding) it shut off. Dreaded debug code of 00 and the CPU led on the asus board indicated failure. I didn't believe it and replaced the mb and same deal.

    I've never had a CPU fail. Other than running the memory at the xmp profile, this wasn't an overclocked system. Am I crazy in thinking CPU's never fail? Especially after a three week break-in and no OC?

    Sigh.
     
  2. Shintai

    Shintai 2[H]4U

    Messages:
    3,692
    Joined:
    Jul 1, 2016
    Asus tends to have auto overclocking on as default. But it shouldn't kill the CPU.

    Contact Intel for RMA, they tend to be very fast in the matters.
     
  3. ManofGod

    ManofGod [H]ardness Supreme

    Messages:
    7,267
    Joined:
    Oct 4, 2007
    It happens, unfortunately. In fact, in regards to the 6700k, when they fail, it is not uncommon for the memory controller to be faulty on it. (I had one that it occurred on and others have as well.) Try using just one stick of ram in each slot to see what happens.
     
    auntjemima likes this.
  4. stormslayer

    stormslayer [H]Lite

    Messages:
    69
    Joined:
    Apr 26, 2005
    Thanks everyone for responding. I tried one stick of RAM and that got nowhere (on two different mb's). In fact, I tried NO RAM at all just to see if I could at least get a different debug code than the dreaded 00 and no luck. Does anyone know how accurate that little LED (near the large power connector to the mb is) for CPU health is on asus boards? Asus has what they call "Q led's" for the cpu, RAM, vga, and boot drive.
    I contacted Intel for an RMA so hopefully this is resolved quickly.
     
  5. Spoonie_G

    Spoonie_G Limp Gawd

    Messages:
    217
    Joined:
    Jan 10, 2007
    PC builds seem to be getting worse and not better. My Z170 build was a nightmare. Enough so for me to never build a PC again (I've been building PC's since 1994). I never want to go through that ever again.
     
  6. OrangeKhrush

    OrangeKhrush Gawd

    Messages:
    530
    Joined:
    Dec 15, 2016
    You guys should try an ITX build in a coolermaster 120 Elite, it is like a 5000 piece puzzle of frustration
     
    Doc Doc likes this.
  7. Emission

    Emission 2[H]4U

    Messages:
    3,448
    Joined:
    Dec 6, 2005
    They're pretty spot on. We used to build systems with the high end ASUS boards back at my old job all the time, and 00 can mean a few different things but they're all related to the boards inability to initialize the CPU successfully (CPU power not plugged in, certain pins bent in the socket, dead CPU).
     
  8. Sufu

    Sufu [H]ard|Gawd

    Messages:
    1,884
    Joined:
    Sep 3, 2006
    That's strange because I had the same exact issue when I built my new system. 6900k on an Asus x99 ROG Strixx. PC shut off randomly, and I got the 00 code. First I figured the MB died, but an RMA later and same issue, I assumed the MB mustoff killed the CPU. First time I've ever had to RMA a CPU, in well over a decade of computing.
     
  9. stormslayer

    stormslayer [H]Lite

    Messages:
    69
    Joined:
    Apr 26, 2005
    Huh. I know this is an N of 2, but do we blame Asus or Intel?
     
  10. Nenu

    Nenu Pick your own.....you deserve it.

    Messages:
    16,965
    Joined:
    Apr 28, 2007
    You dont blame, shit happens.
    No mfr has 100% QA.
     
    pendragon1 and Shintai like this.
  11. Spoonie_G

    Spoonie_G Limp Gawd

    Messages:
    217
    Joined:
    Jan 10, 2007
    You pay money, its supposed to work. Don't tell me or anyone else how we should react when we see our earned dollars being spent on something that doesn't work. Just because you are cool with spending money on stuff that doesn't work don't expect everyone else to be.
     
    Copyright likes this.
  12. WillC310

    WillC310 Limp Gawd

    Messages:
    336
    Joined:
    Apr 1, 2010
    CPUs fail but from my experience at a fairly low rate based on my personal experience over 30 years of messing with PCs. I've had server CPUs fail and the replacement part be DOA.
     
  13. Nenu

    Nenu Pick your own.....you deserve it.

    Messages:
    16,965
    Joined:
    Apr 28, 2007
    You would have a point of you had to throw it away when there is a failure.
     
    Hellstoaster and pendragon1 like this.
  14. NoOther

    NoOther [H]ardness Supreme

    Messages:
    4,145
    Joined:
    May 14, 2008
    Invalid. That is what warranties and service are for. Not everything is always 100%, nor is it expected to be. Should you expect your product to work? Yes, but if for some reason it doesn't, you request a replacement or refund. You don't stupidly go railing off at the company because 1 chip out of millions produced ends up bad, especially when you don't even know exactly why it happened. The end result could be you didn't provide the proper power to the MB, or a spike occurred, or a contaminant got in and changed some resistance, or ESD occurred, etc. etc.
     
    Hellstoaster and pendragon1 like this.
  15. Shintai

    Shintai 2[H]4U

    Messages:
    3,692
    Joined:
    Jul 1, 2016
    Yep, if CPUs never failed there would be no need for warranty :)

    The failure rate however is most likely the one of smallest of any regular consumer electronics out there with 0.x% But 0.x% still happens for someone.
     
  16. pendragon1

    pendragon1 [H]ardness Supreme

    Messages:
    4,688
    Joined:
    Oct 7, 2000
    really?! don't tell us we shouldn't be ok with it!
    it has a warranty, get it swapped. it not like youre out the money. not saying how you should react, just that in my opinion you are over reacting.
     
  17. Guarana [BAWLS]

    Guarana [BAWLS] Bad in Bed

    Messages:
    995
    Joined:
    Oct 3, 2001
    I take it you've never worked with anything built before, then?

    Parts go bad. Even with testing, parts break.

    There is no way to guarantee a 0 percent failure rate, it's actually impossible.
     
  18. HeavensCloud

    HeavensCloud Oswego, not shitty as Buffalo

    Messages:
    7,908
    Joined:
    Feb 7, 2008
    I just built an HTPC in one of those, besides cramming the PSU wires in place it wasn't bad at all.
     
  19. Spoonie_G

    Spoonie_G Limp Gawd

    Messages:
    217
    Joined:
    Jan 10, 2007
    I have and when crap breaks I let them know about it.

    Brand new parts? And that's acceptable to you? Testing? What are you testing? It should have already been tested by the manufacturer. The end user isn't an Alpha tester.

    Who's talking about a zero percent failure rate? 1 in 10,000 is acceptable. 100 in 10,000 isn't.. No?
     
  20. Spoonie_G

    Spoonie_G Limp Gawd

    Messages:
    217
    Joined:
    Jan 10, 2007

    Wanting something to work as advertised is over reacting? Something that you spend your hard earned cash for no less. RMA and warranty repair isn't a consumer experience that I enjoy.
     
  21. Rvenger

    Rvenger [H]ard|Gawd

    Messages:
    1,150
    Joined:
    Sep 12, 2012
    Which version Mobo is that? The second gen X99 or first gen X99? Asus had BIOS issues with overvolting randomly in the not so distant past.
     
  22. pendragon1

    pendragon1 [H]ardness Supreme

    Messages:
    4,688
    Joined:
    Oct 7, 2000
    yes because everyone knows(or should) that not every single thing gets tested. they only test a certain amount out of each batch.
    testing?! what testing!? the testing you are supposed to do when you assemble a system to make sure that there at no manufacturer defects! every part of every system I have ever built has gotten at least a couple hours of testing to ensure there are no bad parts. test ram, cpu and gpu for several hours each and then all together. even test NICs with a constant file transfer for a couple hours. I'm assuming you've never built professionally...
     
    Hellstoaster likes this.
  23. Guarana [BAWLS]

    Guarana [BAWLS] Bad in Bed

    Messages:
    995
    Joined:
    Oct 3, 2001
    Well, here's the flaw in your logic:

    If you want parts at a reasonable price, you're going to get machine produced, and boxed, and shipped. Little to no testing.

    If you want every component you get to be stress tested, you're going to add a SIGNIFICANT premium to the cost of every single thing you buy. Most people aren't into that.

    Also the failure rate of most computer components (especially processors) is well under 1%.
     
    Hellstoaster, NoOther and pendragon1 like this.
  24. pendragon1

    pendragon1 [H]ardness Supreme

    Messages:
    4,688
    Joined:
    Oct 7, 2000
    no wanting something to work out of box is an expectation not an over reaction. getting pissed about getting a bad product is. it happens. as I just stated in my above post. you don't want it to but it can so you test. rma and warranty is for farther into the life of a product. if you get a bad product you exchange it within the first couple weeks and its usually just a straight swap.

    well if you are buying a prebuilt it will have been tested to some extent, big oems get little testing, custom built from smaller companies will usually get at least a good full system test. some, like me test each part and then as a whole system. I went the extra mile without extra charges, some charge more for the extra effort. look at puget systems, you pay a big premium but they test every single little thing and warranty everything out the wazoo! edit: yes you get what you pay for, big time!(meant to add that before posting, oops)
     
    Guarana [BAWLS] likes this.
  25. OrangeKhrush

    OrangeKhrush Gawd

    Messages:
    530
    Joined:
    Dec 15, 2016
    I went fully modular to avoid clutter, but the connections are very tight for MITX.
     
  26. Nenu

    Nenu Pick your own.....you deserve it.

    Messages:
    16,965
    Joined:
    Apr 28, 2007
    Nothing measures up to his standards.
    This isnt the planet he was looking for.
     
    Shintai likes this.
  27. Spoonie_G

    Spoonie_G Limp Gawd

    Messages:
    217
    Joined:
    Jan 10, 2007

    Everything needs to be stress tested by the manufacturer. Are you kidding me? The end user isn't a Beta-tester. Do you know how much money they save because they are allowing the end user to weed out the final bugs that they should have weeded out in the first place? I will never defend any company on the grounds that they cant do enough testing. Crazy talk.
     
  28. pendragon1

    pendragon1 [H]ardness Supreme

    Messages:
    4,688
    Joined:
    Oct 7, 2000
    you obviously don't know how manufactutring works and should quit arguing with those that do.
     
    Last edited: Dec 22, 2016
    Guarana [BAWLS] likes this.
  29. Shintai

    Shintai 2[H]4U

    Messages:
    3,692
    Joined:
    Jul 1, 2016
    Even after leaving testing and the factory it can become damaged during transport and environmental conditions. Its not like the manufactor comes and delivers it to your door.
     
  30. Spoonie_G

    Spoonie_G Limp Gawd

    Messages:
    217
    Joined:
    Jan 10, 2007
    So testing doesn't occur in manufacturing? Where's the QA? Stop making excuses.
     
  31. Spoonie_G

    Spoonie_G Limp Gawd

    Messages:
    217
    Joined:
    Jan 10, 2007
    Finally a response that makes sense. Not sure where all this "Its OK for a manufacturer to release a product that wasn't fully tested" B.S is coming from.
     
  32. pendragon1

    pendragon1 [H]ardness Supreme

    Messages:
    4,688
    Joined:
    Oct 7, 2000
    ok then...
    you clearly aren't comprehending whats been posted and I'm wasting my time.
     
    Hellstoaster and Guarana [BAWLS] like this.
  33. Shintai

    Shintai 2[H]4U

    Messages:
    3,692
    Joined:
    Jul 1, 2016
    Its all about cost isn't it. How long, how much etc should they test it? Electronics as such tends to have a few % failure rate. How much do you think it would cost to remove that? CPUs got some, if not the lowest failure rate of any electronics. There is 100x or more higher chance your mobo kills itself, graphics card goes to bunkers etc.

    [​IMG]
     
    Guarana [BAWLS] likes this.
  34. Guarana [BAWLS]

    Guarana [BAWLS] Bad in Bed

    Messages:
    995
    Joined:
    Oct 3, 2001
    Spoonie is just a bad troll, and not worth your effort folks. Let him live in his bubble.
     
    Shintai likes this.
  35. NoOther

    NoOther [H]ardness Supreme

    Messages:
    4,145
    Joined:
    May 14, 2008
    Do you realize that Intel makes 300-400 Million chips per year? You expect them to stress test a million chips per day? Now who is doing the crazy talk?

    Just as an overview, here is a typical path for chips:

    What they do is they create the design, they make a few runs of the design and do all sorts of enhanced functional and stress testing of the design.
    If it doesn't meet the design goals, they go back, make design changes and test again.
    Once they are happy with that, they do a short run and they test the chips on that run and verify they meet the design specs, through functional and stress testing.
    If they have issues, back to design stage. If not, they go to fab test..
    Here they test the process of printing the chips on silicone.
    Once they are sure the chip design is solid and the fab process is solid, they push them out to mass manufacturing.
    During this process they only do functional and temp tests on the chips as they are being manufactured. They test the chips right on the wafers, this is how they generally bin the chips. Then they separate them, package them and test them again. However, they do not fully stress test them as that would take too much time.
    Also during this process, they may elect to pull a wafer or bin of chips aside at random and do additional testing.

    There is a lot more to it than just these simple steps, and some companies may operate slightly differently, but that is the general process. They also setup additional Quality Control measures and Customer Assistance programs to deal with defective parts that somehow miss the checks during production, or possibly meet some ill fate in packaging and transit.

    If you would like to read more about all the controls and processes Intel puts in place, you can find that here.
     
  36. NoOther

    NoOther [H]ardness Supreme

    Messages:
    4,145
    Joined:
    May 14, 2008
    Back to the OP

    I don't know that you blame anyone, but I would start with getting an RMA for the Intel chip. If you try a second chip and it blows, then its most likely the MB and you will likely have to RMA the MB and then the CPU again.

    Any more updates on your system?
     
  37. Sufu

    Sufu [H]ard|Gawd

    Messages:
    1,884
    Joined:
    Sep 3, 2006
    Not sure about OP, but I'm still using MB #2 with CPU #2. MB #1 did show CPU error code 00, so maybe it was the CPU who died and not the MB taking it out, but I'm just guessing. I never got to test a replacement CPU on the first MB since I RMAd that first.
     
  38. stormslayer

    stormslayer [H]Lite

    Messages:
    69
    Joined:
    Apr 26, 2005
    Whoa. Not everyone is merry this season. I asked who to blame between Asus and Intel, not because I'm intolerant of failure or want to yell at someone, but only because I would like to have an idea if this mb + cpu combo is bad, or one particular part is bad. There's a non-zero risk of failure for all electronics, but it's not the case it's the SAME non-zero risk. I was wondering if anyone had more data than the two cases on this thread or insight into how accurate the q led is on asus boards.

    In answer to above questions, I'm using the second generation Asus Rampage V -- it's the edition 10 model. Both the first board I had and the replacement are listed as ver. 1.0, but the first board came with an older bios and the new board has 1003. I've tried the faulty CPU on both boards and it was in fact faulty on both. Intel sent me a replacement CPU and it works with both boards. I'll send back one MB and the faulty CPU tomorrow. N.b., I filed a ticket with Intel and they ignored it for a week. I tried the chat feature for tech support, it was often "unavailable", but when I did get through to a human they solved the problem / sent out a CPU within 24 hours and it arrived the next day.

    General note: I built a 5960x system about 18 months ago with the msi xpower board. I had the first gen of that board and while it worked, there were perpetual weird issues particularly with cold boots and usb peripherals. I bought out of pocket the second gen of that board and it has worked flawlessly since.

    The asus board seems solid but the usb 2.0 ports are a bit wonky and it took some fiddling on the first board to enable the xmp profile for the ram (I'm running 128 gb though). Still breaking in the new board.

    So that's the complete story. X99 from my perspective seems weird. Between the issues I had with the first gen motherboards and a CPU failing, I'm not sure this is as stable as, say, my x58 platforms, but I may be overreacing to limited data.
     
  39. AbRASiON

    AbRASiON Limp Gawd

    Messages:
    165
    Joined:
    Jul 28, 2004
    Sure ITX builds are a challenge but rewarding in the end, when you get it right and you have this cute little tiny box that can do everything some massive reverberating beast can do.
    If you're super lucky and only modest overclock it's just as quiet to boot.
     
    Shintai likes this.
  40. GothamsReckoning

    GothamsReckoning Limp Gawd

    Messages:
    238
    Joined:
    Dec 22, 2015
    ASUS x99 motherboards are also known to overvolt processors. There are lots of stories of the Rampage V over-volting 5960X's.