Announcement

Collapse
No announcement yet.

CPU Life Expectancy

Collapse
X
Collapse
 
  • Page of 2
  • Filter
  • Time
  • Show
Clear All
new posts
Previous 1 2 template Next
    #1

    CPU Life Expectancy

    I had some thoughts about the real life expectancy of a CPU. Knowing that electromigration and heat plays a large role in their life expectancy, how long in years the cpu is expected to function ok at 100% load? Obviously, considering normal case temperatures, not overclocking, and 24/7 running.
    I can't found data about this in datasheets from the two biggest PC processor manufacturers.
    I think there should be diferences between CPU models and generations, mainly because of diferent fabrication process tecnologies (250nm vs 065nm, for example) that makes transistors and interconnections more small and prone to electromigration more easily. But, electromigration is a real issue in long time or just a myth?

    In your's experiences, how long usually lasts a processor at 100% load?

    Thanks!
    Tags: None
    #2
    Re: CPU Life Expectancy

    As long as the CORE temp is within specification. Like 40C-45C even 50C, then all is golden.

    Honestly, I yet to see a CPU "WEAR" out. Only way they get ruined is overheating far too long, dirty power (badcaps, indeed! or poorly designed board) or excessive overclocking.

    Cheers, Wizard

    Comment

      #3
      Q

      I have heard that the design life of a processor, at stock voltages, is 10 years. High temperatures reduce life, as does increased voltage. Overclocking affects life very little. It's the other things that typically go along with it, like increased voltages and higher temperature, that reduce the life of the processor. There's nothing wrong with simply increasing the clock speed as much as possible at stock voltage and putting a bigger cooler on it. For example, the only difference between an E8400 and an E8600 is the multiplier, price, and possibly the heatsink. Intel will even vary the voltage of various processors over time and over the product line. Some E8400's might take 1.25v stock, while others might take 1.1v. Both are stock voltages, both the same clock speed, with the same heatsink. Is the 1.25v one going to live a shorter life than the 1.1v unit? Is Intel selling factory overclocked processors?

      The waters are muddier than one might think!

      Edit:
      I have had a QX9650 (VID 1.25v) degrade on me. I was running 1.36v through it for several months, at 4.0GHz, with a Vapochill. One day it would no longer do 4.0GHz reliably. It would do 3.8GHz with similar voltages. Several months later, all it will do reliably is 3.6GHz. So now (or the last time I ran it in my desktop) I run it with stock voltages at 3.4GHz to avoid further degradation. Intel's max VID for this processor is 1.3625v, so it should have been safe. Apparently Intel's 45nm process is rather sensitive.
      Last edited by acstech; 01-27-2009, 02:35 PM.
      A man convinced against his will is of the same opinion still.

      Comment

        #4
        Re: CPU Life Expectancy

        Honestly, I don't buy into a processor wearing out. The only way I can see a processor dying is as Wizard said, being overclocked, or with bad caps and motherboards.

        Comment

          #5
          Re: CPU Life Expectancy

          i don't think anybody can really report cpu dying of old age...

          Comment

            #6
            Re: CPU Life Expectancy

            Ya really, there are still old 8088's in computers chugging away for some purpose in shops somewhere.

            I still have a atari 2600, those have 4004's in them and it still works fine.

            Comment

              #7
              Re: CPU Life Expectancy

              acstech is correct.
              Intel builds CPUs for 'at least' 10 years service and that is based on this:


              Actual life is hard to say.
              Heat-up and cool-down cycles (off/on) and load change cycles are probably harder on the internals than running at 100% constantly.
              Different materials expand/contract at different rates so every time you change the heat load you are stressing bonds.

              You have a very interesting theory question but from a practical standpoint the working lifetime is so much longer than the 'useful lifetime' (reaching obsolescence) that it doesn't matter enough to Intel or AMD to do the research.

              .
              Mann-Made Global Warming.
              - We should be more concerned about the Intellectual Climate.
              -
              Be who you are and say what you feel, because those who mind don't matter and those who matter don't mind.
              - Dr Seuss
              -
              You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.
              -

              Comment

                #8
                Re: CPU Life Expectancy

                Originally posted by 370forlife
                I still have a atari 2600, those have 4004's in them and it still works fine.
                6507 actually (cheap version of the 6502)


                Originally posted by acstech
                Some E8400's might take 1.25v stock, while others might take 1.1v. Both are stock voltages, both the same clock speed, with the same heatsink. Is the 1.25v one going to live a shorter life than the 1.1v unit?
                I'd say yes. The lower voltage and/or lower clocked versions of the same chip will tend to last longer and be less sensitive to abuse.

                Is Intel selling factory overclocked processors?
                I guess it's just semantics, but I'd say no. They are the original manufacturer, so they get to decide how fast/what voltage they endorse a given chip to run. If anybody downstream decides to change it, then at that point is when I'd call it overclocking.
                Otherwise all chipmakers would be guilty of "overclocking", and I think that dilutes the meaning of the term.

                But granted underneath it all many different models are actually the same design, and can be used safely at different settings than originally marked.

                Comment

                  #9
                  Re: CPU Life Expectancy

                  It is my understanding that all the processors of a similar technology come from the same wafer. Some test out faster than others from a different location on the wafer. This variation in yield is what makes the different speed ratings.

                  Comment

                    #10
                    Re: CPU Life Expectancy

                    Originally posted by bgavin
                    It is my understanding that all the processors of a similar technology come from the same wafer. Some test out faster than others from a different location on the wafer. This variation in yield is what makes the different speed ratings.
                    I hear that, and also i hear that sometimes, due to market and overproduction, they "re rate" CPU's of higher rating that test OK at desired speed to a lower rating just to sell it because of overproduction.
                    And that's the basis of better overclock in one chip than others, apart from other things of course.

                    I really don't know if is true.

                    Comment

                      #11
                      Re: CPU Life Expectancy

                      bgavin and GL9!

                      You are both correct, that is exactly how it works...
                      And happy happy joy joy for the overclocker that finds those "magic" chips

                      As for life expectancy; I would think it is lower for more recent processors
                      Or rather, that more recent processors on recent architectures are much more sensitive to heat and power spikes
                      There's a reason NASA uses very old processors in space you know
                      "The one who says it cannot be done should never interrupt the one who is doing it."

                      Comment

                        #12
                        Re: CPU Life Expectancy

                        well, those are in cold, and running off the batteries...hehe...

                        but yeah, the thing is they don't need much more cpu power..
                        also they use specialized hardware a lot, and i think they use double or triple safety, so they're essentailly doing anything they can to minimize risk...and that usually means keeping it very conservative...

                        recent chips are much more complex, sure, so they probably are somewaht more sensitive to heat....

                        either way i don't think there's anythong to worry...
                        they become obsolete much sooner than anybody will experience any issues...
                        will be interesting to see how that quick aging is affected by cpus going to more cores rather than more speed...
                        it'll probably mean progression is greatly reduced...i mean how many cores usual user needs anyway...

                        Comment

                          #13
                          Re: CPU Life Expectancy

                          Originally posted by Per Hansson
                          bgavin and GL9!

                          You are both correct, that is exactly how it works...
                          And happy happy joy joy for the overclocker that finds those "magic" chips

                          As for life expectancy; I would think it is lower for more recent processors
                          Or rather, that more recent processors on recent architectures are much more sensitive to heat and power spikes
                          There's a reason NASA uses very old processors in space you know

                          Yes bgavin and GL9 are both correct.
                          I have an old 'Navy Buddy' that went on to work for Intel when he retired from the service. He actually runs the machines that make the waffers and does the first stage of checks and grading. [I was offered the same job but turned it down because I didn't want to spend 12 shifts in a clean room in a poopie suit 4 days a week.]

                          But yes, that is exactly what happens.

                          .
                          Mann-Made Global Warming.
                          - We should be more concerned about the Intellectual Climate.
                          -
                          Be who you are and say what you feel, because those who mind don't matter and those who matter don't mind.
                          - Dr Seuss
                          -
                          You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.
                          -

                          Comment

                            #14
                            Re: CPU Life Expectancy

                            i can tell firsthand that the feared electromigration failures did not happen.
                            have a neighbor with a celeron 300a running at 504 mhz at 2.2v core.
                            its in a mame based arcade game that runs 24/7/365.
                            except the downtime when i recapped it.
                            only problems were when the caps started dieing off.
                            its polyd now.
                            bet it runs another 10 years.

                            Comment

                              #15
                              Re: CPU Life Expectancy

                              Typo:
                              '12 shifts' => '12 HOUR shifts'
                              Mann-Made Global Warming.
                              - We should be more concerned about the Intellectual Climate.
                              -
                              Be who you are and say what you feel, because those who mind don't matter and those who matter don't mind.
                              - Dr Seuss
                              -
                              You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.
                              -

                              Comment

                                #16
                                Re: CPU Life Expectancy

                                Man that's much!

                                And all the time wearing that suit?

                                Anyway, i think they pay good for that work... or i'm wrong?

                                Comment

                                  #17
                                  Re: CPU Life Expectancy

                                  Started at a little less than $50k/year but that was in 2000 and that was just starting.
                                  Benefits were pretty grand too.
                                  But that's not enough for me to spend 12hrs a day in an anti-contamination suit.
                                  I've been in those damned suits. IT SUX!
                                  There are easier ways to make the same or better money.

                                  .
                                  Mann-Made Global Warming.
                                  - We should be more concerned about the Intellectual Climate.
                                  -
                                  Be who you are and say what you feel, because those who mind don't matter and those who matter don't mind.
                                  - Dr Seuss
                                  -
                                  You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.
                                  -

                                  Comment

                                    #18
                                    Re: CPU Life Expectancy

                                    I did a lot of contracting at Intel, but that is all gone now. The local campus was 10,000 strong and down to half that now. All those jobs were outsourced to India. Those affected had to train their Indian replacement, or lose their severance package.

                                    Today, standing 12-hour shifts in a poopie suit for $50k plus grand benefits is very appealing.

                                    My campus (Folsom) was not a FAB plant, so I never got close enough to see the wafer machines in action. We did the design layouts and that sort of stuff. Very interesting to see how it all works. Intel really is an amazing place.

                                    I can't remember ever seeing a processor that died of old age. I have several AMD K6 that are truly ancient, and they all run fine. As KC8 notes above, it's the caps, not the CPU.

                                    Comment

                                      #19
                                      Re: CPU Life Expectancy

                                      Yea, I should add that I agree that CPU's don't die of old age (if it wasn't clear by my post)
                                      CPU's die of power spikes and ripple current
                                      But usually the power circuit on the mobo dies before that happens anyway!

                                      i4004; Ya, NASA uses them because their stuff does not require more power
                                      But problem is even if they wanted more processing power it would be difficult (without a specialized design)
                                      The small transistors of current CPU's have a problem in space; they are too suspectible to the noise caused by the sun in the form of electromagnetic radiation
                                      The stuff used in space is 250nm or larger (250nm is AFAIK cutting edge)
                                      What sits in my system right now is a 45nm CPU, quite the difference!
                                      Stuff is space also have other materials to protect it
                                      "The one who says it cannot be done should never interrupt the one who is doing it."

                                      Comment

                                        #20
                                        Re: CPU Life Expectancy

                                        Originally posted by bgavin
                                        Today, standing 12-hour shifts in a poopie suit for $50k plus grand benefits is very appealing.
                                        I'm 6'3" and those damned suits never fit right.
                                        Stand up too fast in one and my testicles get to say hi to my tonsils.

                                        .
                                        Mann-Made Global Warming.
                                        - We should be more concerned about the Intellectual Climate.
                                        -
                                        Be who you are and say what you feel, because those who mind don't matter and those who matter don't mind.
                                        - Dr Seuss
                                        -
                                        You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.
                                        -

                                        Comment

                                        Previous 1 2 template Next
                                        Working...
                                        X