Announcement

Collapse
No announcement yet.

I don't believe test results of some of the top performers.

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • I don't believe test results of some of the top performers.

    A glaring example is the 3rd place score for the month of July . Looking at the specs I can't believe that it scores over 10,000. In fact there are other 8086K CPUS that also seem way to high. I have an 8700K with all cores overclocked to 51.13. Turbo Boost is disabled. Other than my CPU score which is 20254 I score either at the top of the other categories or in a couple of cases second place. This leads me to believe there is some sort of cheating going on or these folks have discovered some sort of method to increase their scores. Perhaps the fact that the 8086K Turbo Boost is 5Ghz not 4.7Ghz as is the value for the 8700K. If the test is somehow using this to calculate the final performance number I don't think this should be the case. Comments are welcomed.

  • #2
    My results

    Comment


    • #3
      I don't see any obvious evidence of cheating, but there were a few duplicate submissions, which have now been deleted.

      Comment


      • #4
        I can verify that these scores are very much real. I just submitted this one using an 8086K, GTX 1080, Samsung 970 Pro 512, and Corsair 4000 CL19 2x16GB on an ASUS Z370-E: https://www.passmark.com/baselines/V...d=105308355983

        Like most of the people near the top of the list, I've spent a lot of time and effort to make sure this system is running as well as I can get it. Not only that, but it takes a lot of experimenting to figure-out what earns a few extra points here and there to get the score up. Here's the record score for the 8086K, at least for the moment.

        Comment


        • #5
          Originally posted by Spinnacre View Post
          I can verify that these scores are very much real. I just submitted this one using an 8086K, GTX 1080, Samsung 970 Pro 512, and Corsair 4000 CL19 2x16GB on an ASUS Z370-E: https://www.passmark.com/baselines/V...d=105308355983

          Like most of the people near the top of the list, I've spent a lot of time and effort to make sure this system is running as well as I can get it. Not only that, but it takes a lot of experimenting to figure-out what earns a few extra points here and there to get the score up. Here's the record score for the 8086K, at least for the moment.
          This is an old thread, but I'll add a post to summarize the conclusion (after pursuing it myself to find out what's going on).

          Spinnacre's attachments actually demonstrate the OP's point completely.

          While the rest of his scores are within the realm of possibility for a well-tuned, OC'ed system... his Disk Mark score is over 3.5 times what people are getting with the same equipment. I have the exact same NVMe drive (including capacity). It scores between 18k-22k points even after an immediate TRIM operation.

          This is probably why if you click through to his baseline page, it says "baseline has been excluded from average results due to anomalies in the submitted results."

          Comment


          • #6
            The Samsung drives come with the "RAPID" disk caching software (part of the Samsung Magician software package). Which can distort results.
            https://www.passmark.com/forum/perfo...msung-evo-ssds

            Comment


            • #7
              Originally posted by David (PassMark) View Post
              The Samsung drives come with the "RAPID" disk caching software (part of the Samsung Magician software package). Which can distort results.
              https://www.passmark.com/forum/perfo...msung-evo-ssds
              I suspected the same (I have a 850 PRO using RAPID in another computer), and in the course of trying to replicate the phenomenon... I found out that it's actually only available for their SATA SSDs. RAPID isn't available for their NVMe drives, of which the 970 PRO is. It seems like more is going on than just enabling a feature.
              Last edited by DtEW; 10-01-2018, 05:13 AM.

              Comment

              Working...
              X