xXSebaSXx Benching Live Today (Finished)

Add Your Comments

Overclockers Benching Team member xXSebaSXx (HWBot profile) is benching live right now! His “trusty daily driver finally gets put through the ringer.”

Tune in to Overclockers Live! right now for some overclocking with dry ice (DICE). Here’s the low-down:

  • Mobo: Gigabyte GA-MA790FXT-UD5P
  • CPU: AMD Phenom II X4 955BE
  • RAM: Mushkin Redlines (996750) 2x2GB 2000MHz 8-8-7
  • GPU: ATI X300 (”  She’s good enough to feed the monitor so I can see what I’m doing.”)
  • Cooling: Koolance V2 pot + dry ice on the CPU
  • Focus: CPUz Max, SuperPi1M, SuperPi32M, WPrime 32M, WPrime 1024M, PiFast
  • Click here to watch on Overclockers Live!
  • Date & Time: NOW – October 15, 2011 3:20pm 
    • Benching will wind down whenever the DICE runs out, but he currently has about 15lbs left, so there’s a good few hours of hardware torturing to go!

Hope you can stop by root him on!

Jeremy Vaughan (hokiealumnus)

Leave a Reply

Your email address will not be published. Required fields are marked *

Discussion
  1. One of the things for Seeb's WP efficiency is that he's watched and listened to the majority of my streams:sn:

    EDIT:

    Too much eraser will get the ram wet, having "cold walls" of open PCB (no eraser, just paper towels/foam/etc) helps a lot.

    I aim a nidec at around 50% over the ram to keep it warm.
    cold.nut
    good news :)! that is awesome :thup:

    you must tell us your secrets for efficiency :drool:!


    It is very good news. I was more worried about the RAM than anything else; did not want to lose my hypers on the first cold run. :D

    As for efficiency...

    All my runs except the SPi1M were done on microXP. SPi1M was done on tinyVista (I tested microXP with the same settings and Vista came out in front). After I got my Spi1M run is when I started having troubles; first I borked my OSs, then the dual channel issues came about.. So once I got back up and running I just installed microXP and went to town with it.

    microXP is very light out of the door, but it still leaves you some room for improvement. I did my original OS installations on Friday night and went over every single service (using BlackViper's guides) disabling what I could. When it was all said and done I had microXP even lighter than it normally is. I forgot to do a full comparison, but I can tell you this microXP at default settings will have about 10 running services, 1800 Handles and 200ish threads (you can see this in Task Manager in the Performance Tab) and by the time I was done with it; it was running 4 services, 1150 Handles and 120 threads. Physical memory usage went from about 18% to 5% as well.

    So when I re-installed microXP; I re-did all my "services tweaks" by memory and used that for the rest of the session. This OS seems to be quite good for the other 2D benches (WPrime and PiFast). I compared my runs against the runs I was trying to beat over at the Bot and I bested those times, usually on my first try, with slower clocks. So either the guy I was going after (FTW from OCN) was using a terrible OS for his runs (I doubt this as he is an experienced bencher) or we have found the king of OSs for those benchmarks. In all seriuosness, get microXP and give it a try; you won't be disappointed.

    After that it was all the same stuff we have over at the LBB. :D
    xXSebaSXx
    Update:

    Everything survived yesterday's problems. No HW was lost to the moisture issues I had around the memory slots; although I have no idea why I had the problems to begin with since this was one of my most thorough insulation jobs so far. :shock:

    Well... Good news is the board is now powered up and running in dual channel without any trouble at all, my CPU lived to fight another day and, most importantly, my hypers survived!. :D


    good news :)! that is awesome :thup:

    you must tell us your secrets for efficiency :drool:!
    Update:

    Everything survived yesterday's problems. No HW was lost to the moisture issues I had around the memory slots; although I have no idea why I had the problems to begin with since this was one of my most thorough insulation jobs so far. :shock:

    Well... Good news is the board is now powered up and running in dual channel without any trouble at all, my CPU lived to fight another day and, most importantly, my hypers survived!. :D
    Thanks Seba:( I had to work during those hours lol and I wanted to see how you made out on this run... But I see you had a good run and congrats on the run and the numbers you got because all that hard work pay off .. :popcorn::ty::attn::attn::thup:
    I was just looking at my submissions at the Bot and found something that made them even better to me... :D

    Efficiency is a beautiful thing.

    Look at the scores around mine and compare their clocks to mine... :sn:
    The session was good and bad... My board started giving me issues after about two hours under the cold and flat out refused to run dual channel any more. It's not my hypers; I tested them individually and both sticks work fine in single channel, but the board refuses to run dual channel. Even with those problems I managed to get a few good scores.

    Here is tonight's take:

    CPUz Max: 5980.03 -> That's #1 for all 955BEs on DICE. 11.4 Points

    Wprime32: 6.484 -> Again... #1 for all 955BEs on DICE. 13.2 Points

    Wprime1024: 210.656 -> Wait for it... Wait for it... Yup; #1 for all 955BEs on DICE. 7.3 Points

    SPi1M: 11.966 -> This one shows as #2 at the Bot, but #1 shows as running 6.9GHz on DICE... :shock: I don't know if FTW entered the wrong cooling or if he has the golden chip to end all golden chips. I still see it as a personal victory for myself. :clap: 13.8 Points

    PiFast: 19.38 -> Take a wild guess. You guessed it! #1 for all 955BEs on DICE. :comp: 5 Points

    UCBench2011: 1010.6 -> I ran this one just because I had it on the USB drive and it was an easy 2 points to get. Again #1 for all 955BEs, but this time regardless of cooling. :cool: 2 Points

    So all in all 52.7 Points for tonight's session. I did not run SPi32M because it would not have been worth it to run it on single channel memory.

    Thanks to those that dropped by and watched.

    Sebas.