Gaming Comparison, 2gb vs 4gb. 4gb FTW!

Discussion in 'Memory' started by Y3llowbeard, Jun 18, 2008.

  1. Y3llowbeard

    Y3llowbeard Member

    Joined:
    Nov 7, 2006
    Messages:
    654
    Location:
    Atlanta, GA. USA
    One of our guys, Gareth Ogden, did some extensive testing and found some interesting results when gaming with 2gb vs 4gb. One thing a lot of people that read this tend to overlook is the minimum frame rate. If you have gamed on older gear or bought some new title that brings your system to its knees then you'll realize what he's talking about. Max frame rate is a pretty number but, playability is most affected by the minimum framerate and how long a system stays at that minimum frame rate. Who cares if your system can do 500 FPS on an intro screen. When you drop to 10 FPS during a heavy fight you get killed.

    It's sorta long but a good read. http://www.corsairmemory.com/_appnotes/AN804_Gaming_Performance_Analysis.pdf
     
  2. DiGiTaL MoNkEY

    DiGiTaL MoNkEY Inverted Monkey

    Joined:
    Jun 28, 2005
    Messages:
    26,896
    Location:
    Melbourne, Victoria
    A nice little read...but an article written up by a memory company, showing how more memory is better :o
     
  3. OP
    OP
    Y3llowbeard

    Y3llowbeard Member

    Joined:
    Nov 7, 2006
    Messages:
    654
    Location:
    Atlanta, GA. USA
    LOL....well, if we made mice I'd show you an article about how our mouse will help you WIN! :Pirate:

    This was done in response to the question we are seeing literally hundreds of times per day. And, the testing is easily duplicated by any person with the right gear and the inclination to do it. It's actually useful info and not some obscure marketing factoid.
     
  4. Doofus

    Doofus Member

    Joined:
    Jan 31, 2004
    Messages:
    1,738
    Location:
    Wodonga
    Intel would write that one up saying a faster processor is better.
    It's a fairly obvious statement and in the article (Corsair's) they provide a repeatable methodology.

    They could have written something like: Using 4gb instead of 2gb allows for more extereme memory usage with noticeably improved fps in al your favourite games*.

    *Your favourite games may differ from our favourite games and therefore would invalidate the tests by our highly qualified fungineers. Must use Corsair RAM or your computer with run out of the house and steal your silverware.
     
  5. hot_sparx

    hot_sparx (Banned or Deleted)

    Joined:
    May 8, 2007
    Messages:
    299
    I'd say 4gb is better than 2gb. But 6gb rather 4gb it wouldn't really make a difference.
     
  6. Sub.Zer0

    Sub.Zer0 (Banned or Deleted)

    Joined:
    Aug 7, 2007
    Messages:
    3,521
    Location:
    Melbourne
    Yeh it would sound a bit better if it was done by an independent party, I agree with DM.
     
  7. OP
    OP
    Y3llowbeard

    Y3llowbeard Member

    Joined:
    Nov 7, 2006
    Messages:
    654
    Location:
    Atlanta, GA. USA
    Yeah but, that is just the price of fame. :Pirate:

    These tests are easily repeatable if anyone wants to try to prove or disprove the results.
     
  8. Sub.Zer0

    Sub.Zer0 (Banned or Deleted)

    Joined:
    Aug 7, 2007
    Messages:
    3,521
    Location:
    Melbourne
    Nah all good YB. I think you do an awesome job considering some of the bashing you do cop from some of us.
     
  9. Luke212

    Luke212 Member

    Joined:
    Feb 26, 2003
    Messages:
    9,478
    Location:
    Sydney
    The testing is not repeatable. The test fails to mention idle windows memory usage for 2gb or 4gb. They could have loaded up the ram with other junk before testing, to make the results worse.
     
  10. OP
    OP
    Y3llowbeard

    Y3llowbeard Member

    Joined:
    Nov 7, 2006
    Messages:
    654
    Location:
    Atlanta, GA. USA
    That's a bit of an unnecessary low blow. First, my fellow Corsair employee DID NOT do anything to skew the tests. You might want to know what you are talking about before making uninformed comments like that. :thumbdn:

    If you have questions or doubts, just ask. We don't do stuff like you are insinuating.
     
  11. moloko vellocet

    moloko vellocet Member

    Joined:
    Feb 7, 2006
    Messages:
    619
    It can be good to be reluctant to believe a large corporation but it is also worth while looking at all of the facts.

    First being that it's a well known fact that there are bonuses to be obtained from a move from 2GB to 4GB - well all know this. Now at least there's some sort of documented indication as to the reasons why.

    Next up is the guy doing the testing. For those not in the know, prior to working at Corsair, Gareth Ogden was a world wide leader within the IT Journalist trade. He was the founding editor of Custom PC and prior to that was the technical editor for PC Pro - both of these publications are not only respected in the UK but also world wide for the depth and accuracy of the testing.

    I'm just annoyed Gareth beat me to it as I've been pottering away in my spare time looking at 2GB vs 4GB and other memory configurations :lol:

    I'll still continue my pottering away though, as I'll be looking at 2x1GB, 4x1GB, 2x2GB and 4x2GB configurations and am testing on a DDR3 platform as well with more recent chipsets - fingers crossed it doesn't come through feeling like a 'been there, done that' piece :p

    Nicely done Gareth (doubt you'll ever see this post though :p) and well done to Corsair for providing not only enthusiasts but consumers as a whole some more information on potentially expensive purchases.
     
  12. kemd1zl

    kemd1zl (Banned or Deleted)

    Joined:
    Sep 2, 2005
    Messages:
    1,326
    Location:
    Brisbane
    thing is, im fucking awesome. i dont die.
     
  13. OP
    OP
    Y3llowbeard

    Y3llowbeard Member

    Joined:
    Nov 7, 2006
    Messages:
    654
    Location:
    Atlanta, GA. USA
    I'll send him a link as I am certain he would appreciate your comments :thumbup:
     
  14. Luke212

    Luke212 Member

    Joined:
    Feb 26, 2003
    Messages:
    9,478
    Location:
    Sydney
    i didnt say he did or didnt I just said its possible, and he should have made a point of it in the test.
     
  15. Bleemv1

    Bleemv1 Member

    Joined:
    Apr 27, 2007
    Messages:
    1,724
    Location:
    Adelaide
    that's a good review :)
     
  16. OP
    OP
    Y3llowbeard

    Y3llowbeard Member

    Joined:
    Nov 7, 2006
    Messages:
    654
    Location:
    Atlanta, GA. USA
    You clearly stated that the testing is not repeatable and suggested that it is not repeatable because we "could" have done something underhanded to skew the results. The insinuation is unfounded, insulting, and ignorant.

    The test itself is repeatable. If a person repeated the sequence with comparable hardware and got distinctly contradictory results, then you might have grounds to question the methodology. But, the fact is the tests ARE repeatable even if they were run in a different fashion from how you would have done it.

    Since you have an opinion on how this test should have been done, I'm now curious about what your technical credentials are. Done anything worth reading that might contradict the info here?
     
  17. Hugo a gogo

    Hugo a gogo Member

    Joined:
    Jul 7, 2002
    Messages:
    101
    Location:
    Sydney
    My recent experience when upgrading from 2 gig to 4 gig was all good

    I play the BF2 mod Project Reality and on some maps my ram usage (2 gig) would sit on 98% This resulted in write/ read to/from disk that made for a very jerky game.

    I went to 4 gig of ram and although my frame rate didn't really go up what it did do was maintain a higher overall average FPS

    My ram usage now sits on 50 to 60% and I get zero slow downs.

    So +1 for 4 gig of ram when playing games :thumbup:
     
  18. Luke212

    Luke212 Member

    Joined:
    Feb 26, 2003
    Messages:
    9,478
    Location:
    Sydney
    You are rather a strange fellow, and I will restate my suggestion. I suggest that you quote the amount of available ram prior to running the game.
    It is essential for the following reason. If you are using 600MB usage on boot you have 1.4GB to play with. If you are using 300MB on boot, the game will have 1.7GB to play with. If a game uses 1.6GB to play, you can see it will run better in the latter situation, because it will have to swap less.
     
  19. warthogoc

    warthogoc Member

    Joined:
    Dec 27, 2002
    Messages:
    67
    Lol Kemdizl:leet:
     
  20. trance101

    trance101 Member

    Joined:
    Dec 4, 2005
    Messages:
    119
    Location:
    melb
    Nice to see Yb, well before i got the Dominators price was pretty nice with the timmings etc,

    But with the other PQA 1066 4 X 1 gig sticks, it was impossible to get it past a 6400 speed cause u got more sticks to sync thus inviting unstability
    and had a mem bandwith of 8500 mbps


    with the new dominators 2 X 2 gig sticks should be used if running games to get the best timmings and stability, cause less sticks to sync and obviously tested more stringently, thus allowing a mem speed of at least the rated speed or more.

    Mem bandwith went up to pretty close to 10,000 mbps

    but i would think splitting the memory useage between two memory slots, would be more efficient then useing 4 memory slots ( i think this would invite twice as much work createing a slow down) , especially when one uses game applications where the memory useage exceeds 1 gig, thus creating a more smoother game play, especially when useing SLI, because its performance depends alot on CPU and memory performance.

    ie 2X2 gig sticks 1 gig game requirement 500 mb per stick

    ie 4X1 gig sticks 1 gig game requirement 250 mb per stick (2 X times more work in coordinateing mem throughput)

    hmm makes me think of trying a single 2 gig stick to see what happens ?
    Maybe not so noticeable with other applications when useing 4 x 1 gig sticks
     

Share This Page