Results 1 to 10 of 10

Thread: Singlechannel vs. Dualchannel

  1. #1
    IBM
    IBM is offline
    there but for the grace of God, go I IBM's Avatar
    Join Date
    Dec 2003
    Location
    West London
    Posts
    4,187
    Thanks
    149
    Thanked
    244 times in 145 posts
    • IBM's system
      • Motherboard:
      • Asus P5K Deluxe
      • CPU:
      • Intel E6600 Core2Duo 2.40GHz
      • Memory:
      • 2x2GB kit (1GBx2), Ballistix 240-pin DIMM, DDR2 PC2-6400
      • Storage:
      • 150G WD SATA 10k RAPTOR, 500GB WD SATA Enterprise
      • Graphics card(s):
      • Leadtek NVIDIA GeForce PX8800GTS 640MB
      • PSU:
      • CORSAIR HX 620W MODULAR PSU
      • Case:
      • Antec P182 Black Case
      • Monitor(s):
      • Dell 2407WPF A04
      • Internet:
      • domestic zoom

    Singlechannel vs. Dualchannel

    Just looking at the two model of 64bit Athlons, and was wondering if someone might explain something to me. What are the limitations incurred with the two different Integrated Memory Controllers on the two different chips (Athlon64 3200 & FX-51)...The 3200 is singlechannel 64bit while the FX-51 is dualchannel 128bit...

    Any explanations for me?

    Many Thanks
    sig removed by Zak33

  2. #2
    Senior Member
    Join Date
    Sep 2003
    Location
    Guernsey, channel rocks
    Posts
    247
    Thanks
    0
    Thanked
    0 times in 0 posts
    Dual memory controllers effectively means that each module can be written/read to/from at the same time, as a 64-bit memory controller is dedicated to each. This means that on paper you get twice the memory bandwidth, and in reality it does make a difference. On an Abit NF7-M nforce2 board using the onboard geforce4 MX graphics, the score of 4000 marks in 3dmark2001SE was achieved with the RAM modules in the wrong slots.

    Once they were shifted over and DualDDR was enabled, the score jumped to 6000 marks. Of course, this dramatic leap won't be evident in all systems.

  3. #3
    Formerly known as Viet Cong Zombi and tuone
    Join Date
    Aug 2003
    Posts
    1,111
    Thanks
    21
    Thanked
    5 times in 5 posts
    Anyone know any more about this? I remember reading a respected hardware site that affirmed that in all the tests that they had done the difference between single and dual channel performance was in the order of 1-2%, with a whole page devoted to dualchannel in like a 12 page write-up .

    Were the earliest A64 only single channel? As far as I'm sure, Everything 939 is dualchannel. how much performance is gained with dualchannel now that it's integrated? is the difference over 10% In real-world benchmarks?

    Anyone got some wise words on how the on board memory controller responds to overclocking? All I've heard is that the the number one limit on A64 platform is the speed of the RAM.

  4. #4
    Will work for beer... nichomach's Avatar
    Join Date
    Jul 2003
    Location
    Preston, Lancs
    Posts
    6,136
    Thanks
    553
    Thanked
    138 times in 99 posts
    • nichomach's system
      • Motherboard:
      • Gigabyte GA-870A-UD3
      • CPU:
      • AMD Phenom II X6 1055T 95W
      • Memory:
      • 16GB DR3
      • Storage:
      • 1x250GB Maxtor SATAII, 1x 400GB Hitachi SATAII
      • Graphics card(s):
      • Zotac GTX 1060 3GB
      • PSU:
      • Coolermaster 500W
      • Case:
      • Coolermaster Elite 430
      • Operating System:
      • Windows 10
      • Monitor(s):
      • Dell 20" TFT
      • Internet:
      • Virgin Media Cable
    Quote Originally Posted by DaftpuNk
    Of course, this dramatic leap won't be evident in all systems.
    It certainly won't; the onboard graphics on an NF7-M, whilst they are based on a GF4MX core use the systems RAM for memory. Doubling the bandwidth, therefore, will make a HUGE difference in circumstances like that, but significantly less isn systems which are not using integrated graphics dependent upon the systems's RAM.

    I've seen various estimates for the effect of dual-channel memory outside of those cases, and they do vary between 1-5% on overall system performance. The difference will be especially noticeable, however, in any memory intensive tasks.

    Oh, and the above is correct; S754 A64s are single channel only, 939s are dual.

  5. #5
    Senior Member
    Join Date
    Nov 2004
    Posts
    2,028
    Thanks
    3
    Thanked
    34 times in 29 posts
    • YorkieBen's system
      • Motherboard:
      • Gigabyte P35-DS3R
      • CPU:
      • C2D E4600
      • Memory:
      • 4 x 1GB PC6400 Corsair XMS
      • Graphics card(s):
      • Leadtek 8800GT
      • PSU:
      • Corsair 520W
      • Case:
      • Silverstone TJ06
      • Monitor(s):
      • Philips 20" Widescreen
      • Internet:
      • Post Office Broadband
    for A64 systems there isnt a *huge* advantage

  6. #6
    Senior Member
    Join Date
    Feb 2005
    Location
    Liverpool
    Posts
    1,017
    Thanks
    32
    Thanked
    26 times in 20 posts
    • [DW]Cougho's system
      • Motherboard:
      • Asus Crosshair VI Hero
      • CPU:
      • AMD Ryzen 1700 @ 3.9GHz
      • Memory:
      • 16GB Team Group Dark Pro DDR4 @ 3466 C15
      • Storage:
      • 256gb Samsung 830 & 1TB Samsung 850 Evo
      • Graphics card(s):
      • Gigabyte GTX1070 G1 Gaming
      • PSU:
      • Corsair AX760
      • Case:
      • Silverstone FT-05B
      • Operating System:
      • Windows 10
      • Monitor(s):
      • BenQ XL2730Z 1440p 144Hz
      • Internet:
      • BT Infinity 1
    yep, a64's hardly benefit from dual channel.

    As mentioned it ranges from 2% to around 5% better.

    which isnt alot, at all.

    Im happy with my single channel 3400+ anyway.

  7. #7
    Senior Member
    Join Date
    Jan 2005
    Location
    Manchester
    Posts
    2,898
    Thanks
    67
    Thanked
    179 times in 134 posts
    • Butcher's system
      • Motherboard:
      • MSI Z97 Gaming 3
      • CPU:
      • i7-4790K
      • Memory:
      • 8 GB Corsair 1866 MHz
      • Storage:
      • 120GB SSD, 240GB SSD, 2TB HDD
      • Graphics card(s):
      • MSI GTX 970
      • PSU:
      • Antec 650W
      • Case:
      • Big Black Cube!
      • Operating System:
      • Windows 7
    Main issue is dual channel only boosts bandwidth not latency, and latency is what really kills performance with RAM.

  8. #8
    Senior Member
    Join Date
    May 2004
    Location
    Rochester, NY
    Posts
    1,041
    Thanks
    4
    Thanked
    8 times in 8 posts
    • oralpain's system
      • Motherboard:
      • DFI "Blood Iron" P35-T2RL
      • CPU:
      • Intel Pentium E2140 @ 400x8 (3.2GHz), 1.375v
      • Memory:
      • Crucial Ballistix DDR2 800 CL4 @ 500MHz (DDR 1000), 4-4-4-12-T2, 2.3v
      • Storage:
      • 2x Seagate ST3250410AS
      • Graphics card(s):
      • NVIDIA 8800GTS (G92) 512 @ 783MHz core, 1836MHz shader, 1053Mhz memory, stock cooling 70% fan speed
      • PSU:
      • Seasonic SS-500GB
      • Case:
      • Antec P182, with some small modifications
      • Monitor(s):
      • ASUS VW222U
      • Internet:
      • Time Warner "Road Runner" Cable - 16 megabit downstream, 1 megabit upstream
    Bandwidth is more important in a few cases, but normal use and games will see much more benifit from low total latency than from high bandwidth.

  9. #9
    HEXUS webmaster Steve's Avatar
    Join Date
    Nov 2003
    Location
    Bristol
    Posts
    14,269
    Thanks
    286
    Thanked
    829 times in 469 posts
    • Steve's system
      • CPU:
      • Intel i3-350M 2.27GHz
      • Memory:
      • 8GiB Crucial DDR3
      • Storage:
      • 320GB HDD
      • Graphics card(s):
      • Intel HD3000
      • Operating System:
      • Ubuntu 11.10
    Quote Originally Posted by Butcher
    Main issue is dual channel only boosts bandwidth not latency, and latency is what really kills performance with RAM.
    Yeah my CAS0.5 EDO pwns my DDR3200 dual channel
    PHP Code:
    $s = new signature();
    $s->sarcasm()->intellect()->font('Courier New')->display(); 

  10. #10
    Senior Member
    Join Date
    May 2004
    Location
    Rochester, NY
    Posts
    1,041
    Thanks
    4
    Thanked
    8 times in 8 posts
    • oralpain's system
      • Motherboard:
      • DFI "Blood Iron" P35-T2RL
      • CPU:
      • Intel Pentium E2140 @ 400x8 (3.2GHz), 1.375v
      • Memory:
      • Crucial Ballistix DDR2 800 CL4 @ 500MHz (DDR 1000), 4-4-4-12-T2, 2.3v
      • Storage:
      • 2x Seagate ST3250410AS
      • Graphics card(s):
      • NVIDIA 8800GTS (G92) 512 @ 783MHz core, 1836MHz shader, 1053Mhz memory, stock cooling 70% fan speed
      • PSU:
      • Seasonic SS-500GB
      • Case:
      • Antec P182, with some small modifications
      • Monitor(s):
      • ASUS VW222U
      • Internet:
      • Time Warner "Road Runner" Cable - 16 megabit downstream, 1 megabit upstream
    Real latency and memory timings are two very different things, though timings do influence latencies.

    The sytem in my sig has 39ns total latency to main memory. I take this exact ram and run it at the exact same speed/timings in an Athlon XP system and I get well over double that.

    You also need enough badwidth. 60MB/s with 1ns latency is going tobe really really slow for memory. Just as 5000MB/s is going to be slow (for memory) if the latency is in the miliseconds.
    Last edited by oralpain; 17-04-2005 at 04:27 PM.

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
  •