No announcement yet.

Having troubling interpreting memtest results...

  • Filter
  • Time
  • Show
Clear All
new posts

  • Having troubling interpreting memtest results...

    I ran memtest last night with 2 new 8GB RAM modules (on a new linux machine). In the morning, I saw the summary screen and do not understand the results. I think these are the relevant results:
    In the line with Time, Iterations, etc.: Pass: 6 Errors: 0

    So, I thought this meant I have no errors, but below this I have:
    Error Confidence Value: 172,
    Entries for both lower and higher error address,
    Bits in error - total: 32 Min: 1 Max: 31 Avg: 5
    Max Contiguous Errors: 2

    Then, in the Test and Error columns on the right-hand side, each test except 8 has a zero in the error column. Next to test 8 in the error column, the numbers 0-32767 scroll quickly scroll through.

    Any help is greatly appreciated.

  • #2
    Looks very similar to my first set of results (see very recent thread below). It said 'errors 0' and then the additional set of data below with a few 'max contiguous errors', and same massive number on the right scrolling through.

    Not easy to interpret...


    • #3
      mem-noob: I re-ran and got through 9 passes without any errors, so I think I am good. Hope you got yours figured out.


      • #4
        We've confirmed there are some bugs when the error count gets too high. The error count for the individual tests is capped at ~32,000. Once it reaches this it loops back to 0. Additionally if there is a mass of errors in a small section of a test the overall error count won't update until that section of the test is finished, which is slowed down significantly by all the errors. There is also a bug with the 'max continuous errors display', making it unreliable. We'll correct these issues in a future release.

        If there are errors listed for individual tests on the right then there definitely are errors occurring.

        Update: These bugs in MemTest86 were fixed in release V4.2, 19/March/2013


        • #5
          okay, understood. Mine did have errors, then, despite not showing up on my second test. This was already pointed out by David anyway.

          Looks like it's the same for etdoughe because he/ she did have the mass of errors on the right, just like me, during their first test.