Announcement

Collapse
No announcement yet.

I want to go faster (in Seti!)

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

  • #16
    Rags

    Until you tell us how you cheated, I can hardly take it seriously that you complete your WUs in about 1.5 hours on PIII 550 machines.

    I would very much like it if you logged some WU times on the Coppermine and the Katmai (with SetiSpy for instance) and published the logs. That would make for some interesting comparisons.

    Comment


    • #17
      Hurrah!!!

      Finally made a top 20 list

      BTW just finished my fastest WU ever, 5h 27 min

      ------------------
      P3 500, 224 MB ram, G400 16SH, SB Live Value
      Maxtor DM 40+ 30GB, IBM Deskstar 16GP 10GB, Maxtor 4320 13 GB
      6h40min avg on the last 10+ SETI WUs


      [This message has been edited by CHHAS (edited 10 May 2000).]
      "That's right fool! Now I'm a flying talking donkey!"

      P4 2.66, 512 mb PC2700, ATI Radeon 9000, Seagate Barracude IV 80 gb, Acer Al 732 17" TFT

      Comment


      • #18
        Rags,

        It seems there's very little "default" about your configuration ;-) Oh well, if you don't want to share...

        It doesn't however eliminate my suspicion that default Coppermines perform worse than default Katmais at the same speed, and that was what I was hoping for.

        I'd hardly expect to get WU times below two hours from anything short of an Alpha EV6...

        Comment


        • #19
          It gets better and better.

          (I found out that I can just generate HTML-output right from my database. Very nice, and the text-table looks crappy for me, so the weekly update is a little early this time )

          <TABLE BORDER=1><TR><TH>name</TH><TH>wus</TH><TH>hours</TH><TH>average</TH></TR><TR><TD>Rags</TD><TD>34</TD><TD>43.18</TD><TD>1h 16m</TD></TR><TR><TD>Manex</TD><TD>3</TD><TD>8.83</TD><TD>2h 57m</TD></TR><TR><TD>bert</TD><TD>20</TD><TD>87.66</TD><TD>4h 23m</TD></TR><TR><TD>acobra</TD><TD>18</TD><TD>87.66</TD><TD>4h 52m</TD></TR><TR><TD>Mark F.</TD><TD>92</TD><TD>525.96</TD><TD>5h 43m</TD></TR><TR><TD>Eye MD</TD><TD>16</TD><TD>95.55</TD><TD>5h 58m</TD></TR><TR><TD>Jakob Kruse</TD><TD>34</TD><TD>215.09</TD><TD>6h 20m</TD></TR><TR><TD>Jorden</TD><TD>8</TD><TD>50.84</TD><TD>6h 21m</TD></TR><TR><TD>Guyver</TD><TD>41</TD><TD>262.98</TD><TD>6h 25m</TD></TR><TR><TD>Atelier Media</TD><TD>66</TD><TD>438.30</TD><TD>6h 38m</TD></TR><TR><TD>Ant</TD><TD>26</TD><TD>175.32</TD><TD>6h 45m</TD></TR><TR><TD>hominid skull</TD><TD>12</TD><TD>81.63</TD><TD>6h 48m</TD></TR><TR><TD>minsoo</TD><TD>18</TD><TD>123.22</TD><TD>6h 51m</TD></TR><TR><TD>CHHAS</TD><TD>23</TD><TD>159.77</TD><TD>6h 57m</TD></TR><TR><TD>KvHagedorn</TD><TD>41</TD><TD>289.48</TD><TD>7h 4m</TD></TR><TR><TD>VSA</TD><TD>60</TD><TD>438.30</TD><TD>7h 18m</TD></TR><TR><TD>Ackerot</TD><TD>18</TD><TD>132.08</TD><TD>7h 20m</TD></TR><TR><TD>Kruzin</TD><TD>5</TD><TD>36.70</TD><TD>7h 20m</TD></TR><TR><TD>Greebe</TD><TD>59</TD><TD>438.30</TD><TD>7h 26m</TD></TR><TR><TD>Guru</TD><TD>56</TD><TD>422.43</TD><TD>7h 33m</TD></TR></TABLE>

          So, Rags, what do you want us to do before you tell us your secret? Should we get on our knees?

          (Voice from below... ) Won't you please, pretty please tell us how you manage to do your WUs in 1 hour and 16 minutes?

          Martin

          (Edit: wahow! I guess UBB isn't too happy about html-tables...)

          (Re-edit: does this fix it?)


          [This message has been edited by Ees (edited 11 May 2000).]

          Comment


          • #20
            That BIG GAP happens when you get Frontpage/Excel/Word or something like that to produce HTML pages and then cut n paste into this UBB. I can't figure out why though!

            Thos average times - do they mean per WU per CPU? Or say if it takes me 5 hours to do one, and I have 5 machines, is my average time 1 hour? If so, that explains Rags' times.

            ------------------
            Cheers,
            Steve

            "Life is what we make of it, yet most of us just fake"

            Comment


            • #21
              goto http://setiathome.ssl.berkeley.edu/s...ountry_43.html
              and check nr192!!!! 26min....

              MMMMMM im nr79 in Sweden

              Stefan

              Comment


              • #22
                or http://setiathome.ssl.berkeley.edu/stats/venue_1.html
                nr46? 17min

                Comment


                • #23
                  SteveC,

                  Dunno why that big gap appears, but I now know how to fix it: strip all newlines from the table's HTML (make it all one big line).

                  It took some searching before I found the solution (on UBB's own forums).

                  And no: these are really the times per unit.

                  But if I read between the lines correctly, Rags' times are less than official...

                  I apologize up front if I misunderstand, but I sure hope that we can keep Team Matrox Users patch-free.

                  Martin

                  Comment


                  • #24
                    Sort of cheating, Isn’t it

                    Comment


                    • #25
                      Well,
                      check out these guys: <a href="http://setiathome.ssl.berkeley.edu/cgi-bin/cgi?cmd=team_lookup&name=vietnam">Vietnam</a>
                      Sh*t, 3 minutes!

                      Jan M.

                      Comment


                      • #26
                        Rags have you found out a way of using the P3 SSE instructions?

                        ------------------
                        PIII450@558, ABIT BX6-2, 256RAM, G400MAX, SBLIVE, HOTROD-UDMA66
                        Join the MURC SETI team @ <A HREF="http://setiathome.ssl.berkeley.edu/cgi-bin/cgi?cmd=team_join_form&id=25678


                        " TARGET=_blank>http://setiathome.ssl.berkeley.edu/cgi-bin/cgi?cmd=team_join_form&id=25678


                        </A>
                        According to the latest official figures, 43% of all statistics are totally worthless...

                        Comment


                        • #27
                          Rags has is now gone from the SetiMURC team please make him come back

                          ------------------
                          PIII450@558, ABIT BX6-2, 256RAM, G400MAX, SBLIVE, HOTROD-UDMA66
                          Join the MURC SETI team @ <A HREF="http://setiathome.ssl.berkeley.edu/cgi-bin/cgi?cmd=team_join_form&id=25678


                          " TARGET=_blank>http://setiathome.ssl.berkeley.edu/cgi-bin/cgi?cmd=team_join_form&id=25678


                          </A>
                          According to the latest official figures, 43% of all statistics are totally worthless...

                          Comment


                          • #28
                            Oh, look at that, I'm running the same times as Mark F. and Greebe, with considerable less WU's

                            Matt, did you have a couple of 1 minute WU's in there?

                            Jord.

                            [This message has been edited by Jorden (edited 11 May 2000).]
                            Jordâ„¢

                            Comment


                            • #29
                              Rags,
                              there is no way that you could achieve those times with a P3. The Compaq Washington benchmark centre are averaging 1 hour 39 minutes, and they probably have some pretty funky machines. So what are you running?
                              Jan M.

                              Comment


                              • #30
                                Sorry I cant make him come back Maybe my WU may help U....


                                Hm... I have now completed 130 WU...
                                Why is it showing 18...
                                ---------------------------
                                Nice my P!!!450@600 isn't so slow
                                Running the GUI 2.04 client...
                                "Ackerot 130 1018h 25m..."

                                [This message has been edited by Mr D. Ackerot (edited 12 May 2000).]

                                Comment

                                Working...
                                X