Little to none benefit with dual or tripple channel memory?

Discussion in 'PC Industry' started by MistaPi, Jul 14, 2010.

  1. MistaPi

    Regular

    Joined:
    Jun 12, 2002
    Messages:
    362
    Likes Received:
    7
    Location:
    Norway
  2. Davros

    Legend

    Joined:
    Jun 7, 2004
    Messages:
    15,504
    Likes Received:
    2,684
    pcmark vantage isnt enough to draw a conclusion from
     
  3. MistaPi

    Regular

    Joined:
    Jun 12, 2002
    Messages:
    362
    Likes Received:
    7
    Location:
    Norway
    Fair enough. But is there any real world application benchmarks that show a big gain with dual-channel memory, with say Corei5/i7 or Phenom ii? I've tried google it without much luck.
     
  4. Silent_Buddha

    Legend

    Joined:
    Mar 13, 2007
    Messages:
    16,871
    Likes Received:
    6,032
    No, there's virtually no difference in most consumer applications and in most games. There's some sites that have tested Lynnfield (double channel) versus Bloomfield (triple channel) and clock for clock the difference is negligible in almost all cases. Granted there's a few architectural changes between the chips also, but those shouldn't result in any large performance disparities.

    Regards,
    SB
     
  5. Davros

    Legend

    Joined:
    Jun 7, 2004
    Messages:
    15,504
    Likes Received:
    2,684
    the main advantage of getting a triple channel board is they have more ddr slots (if you need a lot of ram of course)
     
  6. neliz

    neliz GIGABYTE Man
    Veteran

    Joined:
    Mar 30, 2005
    Messages:
    4,904
    Likes Received:
    23
    Location:
    In the know
  7. fellix

    fellix Hey, You!
    Veteran

    Joined:
    Dec 4, 2004
    Messages:
    3,503
    Likes Received:
    420
    Location:
    Varna, Bulgaria
    Bloomfield got its triple channel interface because it was meant primary as a server MPU -- not much for desktop, really. The wider DDR3 interface was put to balance out the ultra-aggressive hardware prefetch mechanism in Nehalem's architecture, in combination with its fast and big L3 cache. Running out eight threads in parallel can generate many outstanding I/O requests in a typical server environment, where most of the tasks are I/O bound, anyway. Some admins had to turn off the HW prefetcher in Core 2 based Xeon systems, because it was way too aggressive (not to match Nehalem, though) for its tiny FSB interface, or see reduced overall performance.
    And yes, the other reason is that with three channels it can simply handle more DIMMs at higher clock-rate. ;)
     
  8. metafor

    Regular

    Joined:
    May 26, 2010
    Messages:
    463
    Likes Received:
    0
    On top of that, Quickpath allows one processor to directly fill its cache with the memory pool of another. So each memory controller not only has to provide data for its local processor but potentially data for others.
     
  9. BRiT

    BRiT Verified (╯°□°)╯
    Moderator Legend Alpha

    Joined:
    Feb 7, 2002
    Messages:
    14,202
    Likes Received:
    11,767
    Location:
    Cleveland
    Overclockers who are doing 'suicide runs' only use single DIMMS. They're going for highest overclock, not best performance. This maximizes their stability because it minimizes power draw and stress load on the CPU and Motherboards.
     
Loading...

Share This Page

  • About Us

    Beyond3D has been around for over a decade and prides itself on being the best place on the web for in-depth, technically-driven discussion and analysis of 3D graphics hardware. If you love pixels and transistors, you've come to the right place!

    Beyond3D is proudly published by GPU Tools Ltd.
Loading...