Barton Posted June 25, 2011 Share Posted June 25, 2011 Sweet, it's not that one run is bugged. Mr Paco has indicated that these runs are consistently repeatable. Virtually every PCMark 05 benchmark session on his 8 i-ram set-up with that processor ends with high total PCMark 05 scores and high sub-test scores. Without access to the PCMark 05 algorithm for the final score and for the sub-test scores we can't explain the internal mechanics of how the benchmark program computes those scores other than to say everything is on the up and up, and run totally within the HWBot rules for PCMark 05. I understand where Christian Ney is coming from with his initial comments. All we can say is that the extremely fast hard drive test times attributable to the i-rams must also affect favorably various other sub-test scores in XP. If the results are repeatable time and time again, it's not a "bugged" run. Rather it is the result of extremely fast hardware and a great deal of skill on the part of Mr.Paco in putting that build together and getting it to run that fast. HWBot doesn't prohibit fast hardware, it encourages that. Please mention the above points to the other members of the staff and see about approving the run. There appears to be no valid reason to deny it. It's not bugged. It's just darn fast. ;>) Quote Link to comment Share on other sites More sharing options...
xoqolatl Posted June 25, 2011 Share Posted June 25, 2011 Without access to the PCMark 05 algorithm for the final score and for the sub-test scores we can't explain the internal mechanics of how the benchmark program computes those scores But we do have access to that, read the PCMark05 white paper. Quote Link to comment Share on other sites More sharing options...
Barton Posted June 25, 2011 Share Posted June 25, 2011 The white paper says on page 22 under limitations essentially that individual component test scores are affected by other component test results. That's the point I mention in the post above. What we don't know, and what the white paper does not seem to tell us, is the exact mathematical amount of that effect. Quote Link to comment Share on other sites More sharing options...
Christian Ney Posted June 25, 2011 Share Posted June 25, 2011 You mean that ? 87 [(Test 1 x Test 2 x Test 3 etc) ^(1/16)] = PCMark05 Score Quote Link to comment Share on other sites More sharing options...
Christian Ney Posted June 25, 2011 Share Posted June 25, 2011 After investigation by hwbot staff, there were no cheats involved nor illegal tweaks. But unfortunately mrpaco was the victim of a know bug with Pentium 4 521 + PCMark05 that cause the ''Multithreaded test #1''(aka audio compression + video encoding) to bug and give a too high score. Indeed, we have been already confronted to the same suituation 6 months ago with another Pentium 4 520 having a score of 18k without any IRams and with an out of line ''Multithreaded test #1'' score. After we asked him to run it with a lower fsb he told us that with 1 less MHz on the fsb he had a score of 11k and all tests were now right accoring to the cpu frequency and the hardware used. We are sure that if mrpaco rerun PCMark05 with 1 or 2 less MHz on the FSB he will have a non bugged run that will still be a very good result according to his hardware and skills. We encourage him, because he can do a huge 12~13k PCMark05 score and maybe even more and we also want to thank him for his total cooperation. Quote Link to comment Share on other sites More sharing options...
mr.paco Posted June 25, 2011 Share Posted June 25, 2011 Thank you to all Bot staff for the time you put into examining my situation and I can respect the reason for the final word. Not that I like it . Would have been nice to have been able to keep it, but it is what it is. Thank you to everyone for your support; and thank you to the nay sayers you have given me yet one more reason why to push even harder Again thank you... Now where is that 521??? LOL Quote Link to comment Share on other sites More sharing options...
Christian Ney Posted June 25, 2011 Share Posted June 25, 2011 I know you are able to do at least 13K But unfortunately this 18k and the 25k you provided us are bugged Quote Link to comment Share on other sites More sharing options...
Mr.Scott Posted June 25, 2011 Author Share Posted June 25, 2011 Link to the other bugged run please.......if it hasn't been removed. Also, thanks to the staff for taking the time to look and not just blowing it out of the water. And a personal thanks to Alex.........just because I like Alex and he's always been straight up with me. Quote Link to comment Share on other sites More sharing options...
Christian Ney Posted June 25, 2011 Share Posted June 25, 2011 Link to the other bugged run please.......if it hasn't been removed. Submission: http://hwbot.org/forum/showthread.php?t=11904 Score: http://3dmark.com/pcm05/2344343 Quote Link to comment Share on other sites More sharing options...
Mr.Scott Posted June 25, 2011 Author Share Posted June 25, 2011 Thank you CN. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.