Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 5 of 7 FirstFirst ... 34567 LastLast
Results 41 to 50 of 61

Thread: nForce2 Dual Channel Question




  1. #41
    Join Date
    Dec 2002
    Posts
    4,246

    Default

    Alright, this is enough of this crap. You clearly don't get the point. If you are so unhappy here, then why do you stay?
    I've gone too far and need to move on!

  2. #42
    Beefy Guest

    Default

    OK, after actually reading over this thread, I'm actually gonna re-open it (as long as everyone keeps calm...).

    Now I'm definitely no expert when it comes to overclocking or technical stuff about hardware, but I know what I see when I use my computer... I originally ran my PC from 1 stick of RAM (Soltek nForce2 MB) until my Twin Corsair stuff turned up. I didn't change timings, I wasn't overclocking, but I did notice a performance increase when running dual channel. I don't totally rely on benchmarks to verify performance. As long as I can notice a change, then I'm happy.

    When I went from single to dual, everything just seemed quicker. The little blue bar at the start of Windows didn't do as many laps, the welcome screen now flashes up then off again. I never benchmarked the differences between the two, and didn't really care.

    Now I'm definitely not saying that the benchmarks are false, but I'm less inclined to believe them if they only say a 2.5% difference when I can see things happening quicker...

    However, all in all, there's no point bickering about this... Can we get some decent ideas as to why people believe what they do?

  3. #43
    Join Date
    Jan 2003
    Posts
    968

    Default

    Yeah, despite the bickering this has been a good thread. The original question was answered, and I personally find it quite interesting that certain motherboards can possibly run in DC mode with 3 identical sticks of RAM. At one point when one of my machines was down, I ran a third stick of Corsair PC3500 XMS on my A7N8X....I didn't do benchies, but it didn't seem to make much difference one way or another. When I have some time, I'll do my own tests....just out of curiosity. :cheers:
    <img src="http://www.statgfx.com/statgfx/folding/?&username=Wordbiker&border=0,255,0&tr ans=yes&template=fah_original&.jpg" alt="www.Statgfx.com" />
    Revenge is just a respawn away....Pagosa LAN Parties

  4. #44
    Beefy Guest

    Default

    OK, I just had a little hunt around, and found a couple single vs dual benchmarks. The best one was on the following page - http://www.digit-life.com/articles2/...-1vs2channels/

    Right down the bottom of the page is a very handy little table that shows the percentage increase with dual channel memory. Now it shows that while a lot of things don't have that much of a performance increase, some things actually do make a decent difference (ignoring the IGP part of the table).

    Then again, how many of these tests really do push memory to it's limits?

  5. #45
    Join Date
    Nov 2001
    Location
    New England Highlands, Australia
    Posts
    21,907

    Default

    I'm a multi-taskin' addict runnin' with at least SETI@home, WinAMP, Trillian, 3 or 4 IE windows, PSP, calculator plus whatever else happens to be needed doin' at the time (rippin', burnin', gamin', etc) but with only one channel operatin' all this becomes a chore in waitin' due to slowness though with both channels in operation everything runs quickly and smoothly. Benchtests are all very good to give ppl an idea on a startin' point but real world usage is what really talks to me.

  6. #46
    Beefy Guest

    Default

    Same here. I use WinAMP 3, which a lot of people complain and say is 'resource hungry'... yet I don't notice a thing. Also, running all my All-in-Wonder software, as well as my Audigy 2 software... and running everything else, that extra bit of oomph in the memory area helps.

  7. #47
    Join Date
    Sep 2003
    Posts
    184

    Default

    Beefy: That article came up with roughly the same numbers as I did. 2-4%. Whether or not windows does in fact load faster for you, or it just seems like it, well I guess the only way to test that is by doing a more accurate test with a stopwatch.

    As far as the tests run go, I'd say they definitely stress the memory. MPEG4 encoding loves memory bandwidth, but like I tested there's only a ~2.7% increase in bandwidth using dual channel, which is exactly how much faster the encoding is.

    I think the gain from having two memory controllers is that it can access and retrieve two pieces of data (assuming they're stored on different sticks) at the same time and have them both READY to go to the CPU. Since the cpu memory bus is only 64 bits wide, which is how wide the memory controllers are each, only 1 piece of data can be sent through. The small gain though is when 1 piece of data is sent and the other is already there waiting thanks to a second memory controller, instead of the first one having to go back and forth again. Judging from the tests though that portion obviously doesn't take long at all. Barely long enough to make a marginal difference in most scenarios.

    Just for yucks I've run in single channel for the last day or so to see if I noticed a difference, but I can't say I have. Even in games I don't really notice any FPS drops or slower loading times. Windows loads in the exact same number of bar passes. For me it's definitely not noticeable, but like it's been shown running in dual channel is basically an extra 2-3% gain for free, so why not utilize it.

  8. #48
    Join Date
    Sep 2003
    Posts
    33

    Default

    Alot depends on whether you run your system overclocked or not. If you are running stock 133 or 166 FSB then Dual channel can make upwards of 50% or more bandwidth. This is significant:eek: Unfortunately if you are OCing then at 200fsb (maybe lower) it only makes 2% to 6% difference. The Increased performance of running 200FSB or better more than makes up for the difference in loss of DC advantage. 2% to 6% difference IS a difference and closly approximates the cost increase for DC so that is a wash when OCing. There is no reason not to use DC except that it often limits your top OC fsb. It is definately excellent when running your sytem near or at stock speeds on all but two of AMD current CPUs.

    Hope this clears up why there are two completely opposing veiw on this subject. I promise that my info is spot on. I make a rule of posting accurate info.

    BTW I was never rude in this forum although I did call somebody "just a reader" after they COMPLETELY discounted my post like I was a moron. No hard feelings in my case though:)
    3544mhz 295FSB
    2.4C P4C800-E
    KHXPC3200 234mhz
    22254 Ghettocooled
    22532 3DM01

  9. #49
    Join Date
    Feb 2003
    Location
    Fincastle, IN, USA
    Posts
    3,776

    Default

    Im just curious..what doe Ghettocooled in ya sig mean?

  10. #50
    Join Date
    Sep 2003
    Posts
    33

    Default

    Cold air cooling. I have a $100US airconditioning unit as a casefan and a semidecent copperheatpipe HS/60mm 7000rpm fan. CPU temps are about 20C at 2.01v;)
    3544mhz 295FSB
    2.4C P4C800-E
    KHXPC3200 234mhz
    22254 Ghettocooled
    22532 3DM01

Page 5 of 7 FirstFirst ... 34567 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •