1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Best memory benchmarks... 64bit

Discussion in 'Motherboards & Memory' started by Sasqui, Nov 21, 2012.

  1. Sasqui

    Sasqui

    Joined:
    Dec 6, 2005
    Messages:
    8,728 (2.17/day)
    Thanks Received:
    2,713
    Location:
    Manchester, NH
    What is a free, and good, system memory benchmark? I downloaded AIDA64, but the trial version is hobbled.
     
    10 Year Member at TPU Crunching for Team TPU
  2. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    15,770 (4.05/day)
    Thanks Received:
    9,572
    Location:
    Beaumont, Alberta
    AIDA64, Maxxmem, and SiSoft Sandra.


    AID64 is a worthy purchase, BTW. Service is top-notch, really.
     
    10 Year Member at TPU
  3. natr0n

    natr0n

    Joined:
    Jan 29, 2012
    Messages:
    3,522 (1.99/day)
    Thanks Received:
    2,403
    Location:
    Boca Raton, Florida
    AIDA64 is good. Maxxmem shows higher numbers. I used to like sandra back in the day now its confusing.
     
  4. 95Viper

    95Viper

    Joined:
    Oct 12, 2008
    Messages:
    5,524 (1.86/day)
    Thanks Received:
    2,526
    Location:
    στο άλφα έως ωμέγα
  5. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    15,770 (4.05/day)
    Thanks Received:
    9,572
    Location:
    Beaumont, Alberta
    Meh, Sandra has always been that way, IMHO. Gives very different numbers from the rest, but does include CPU cache testing too, and since this is affected by ram speed too, I use Sandra in my memory reviews.
     
    Sasqui and 95Viper say thanks.
    10 Year Member at TPU
  6. 95Viper

    95Viper

    Joined:
    Oct 12, 2008
    Messages:
    5,524 (1.86/day)
    Thanks Received:
    2,526
    Location:
    στο άλφα έως ωμέγα
    +1
    Yep, it seems to the standard everyone uses.
     
  7. LAN_deRf_HA

    LAN_deRf_HA

    Joined:
    Apr 4, 2008
    Messages:
    4,647 (1.47/day)
    Thanks Received:
    994
  8. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    15,770 (4.05/day)
    Thanks Received:
    9,572
    Location:
    Beaumont, Alberta
    SupCom, and Shogun2's CPU benchmark. Civ5 too, a bit.

    Then there's 3dm01...

    But I think he was looking for something that displays bandwidth, although I could be wrong.

    Codemasters titles seem to be memory sensitive as well, although...it's more CPU cache, it seems? not exactly sure.
     
    10 Year Member at TPU
  9. Sasqui

    Sasqui

    Joined:
    Dec 6, 2005
    Messages:
    8,728 (2.17/day)
    Thanks Received:
    2,713
    Location:
    Manchester, NH
    Is it a free, non-hobbled trial? I just ran across PassMark, any good?
     
    10 Year Member at TPU Crunching for Team TPU
  10. Completely Bonkers New Member

    Joined:
    Feb 6, 2007
    Messages:
    2,579 (0.72/day)
    Thanks Received:
    516
    Sasqui says thanks.
  11. Sasqui

    Sasqui

    Joined:
    Dec 6, 2005
    Messages:
    8,728 (2.17/day)
    Thanks Received:
    2,713
    Location:
    Manchester, NH
    10 Year Member at TPU Crunching for Team TPU
  12. DRDNA

    DRDNA

    Joined:
    Feb 19, 2006
    Messages:
    4,844 (1.23/day)
    Thanks Received:
    598
    Location:
    New York
    Last edited: Nov 21, 2012
    Sasqui says thanks.
    10 Year Member at TPU
  13. Sasqui

    Sasqui

    Joined:
    Dec 6, 2005
    Messages:
    8,728 (2.17/day)
    Thanks Received:
    2,713
    Location:
    Manchester, NH
    So what I'm trying to accomplish:

    Bench between different memory timings, including testing between 1T and 2T command rates, memory speeds and CAS timings. I expecially want to see raw write and read speeds. According to Passmark, I'm getting 24 Gb/s, does that sound right for 2133 CAS9 ram?

    Stress tests to determin stability.

    Will SuperPi do that?
     
    10 Year Member at TPU Crunching for Team TPU
  14. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    15,770 (4.05/day)
    Thanks Received:
    9,572
    Location:
    Beaumont, Alberta
    SPi for stability, yes, run 32m. if it passes, it's usually pretty close, anyway.

    Also good for checking the gains or losses form minor timing changes, BTW.

    Perforamnce given relates to nothing, beucase it's X87, but X87 really hammer cache and IMC on Intel chips. Newest AMD chips just use software x87 emulation, AFAIK.
     
    10 Year Member at TPU
  15. drdeathx

    drdeathx

    Joined:
    May 14, 2009
    Messages:
    2,132 (0.77/day)
    Thanks Received:
    481
    Location:
    Chicago burbs
    SuperPi is outdated and single threaded. Pretty much a fail benchmark by todays standards.
     
  16. Tatty_One

    Tatty_One Super Moderator Staff Member

    Joined:
    Jan 18, 2006
    Messages:
    18,818 (4.74/day)
    Thanks Received:
    4,736
    Location:
    Worcestershire, UK
    cadaveca says thanks.
    10 Year Member at TPU
  17. cadaveca

    cadaveca My name is Dave

    Joined:
    Apr 10, 2006
    Messages:
    15,770 (4.05/day)
    Thanks Received:
    9,572
    Location:
    Beaumont, Alberta
    meh, if you cannot see how to use it, I'd call that a fail in comprehension. It's not meant to judge actual performance at this point, for sure. However, it is awesome for stability testing and ensuring timing changes do what they are supposed to. That is all.

    final benchmark score is determined by CPU speed as well, so I definitely agree on it having limited usefulness, but it's not to be ignored. Ignoring something means you are ignorant of it...:p
     
    10 Year Member at TPU

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)