nForce2 Dual Channel Question

[Wordbiker]

Quote from my A7N8X DX user's manual:
NOTE: To enhance system performance, utilize dual-channel feature when installing additional DIMMs. You may install the DIMMs in the following sequence: Sockets 1 and 3, sockets 2 and 3 or sockets 1, 2 and 3.
Hmmm, doesn't say anything about losing dual-channel with a 3rd DIMM....
My Albatron KM18G Pro manual states:
To enable dual-channel DDR functionality, your system must meet certain configuration conditions. First, you can only use 2 of the 3 DIMM slots...
And my Chaintech 7NIF2 Summit manual notes:
For maximized Dual-Channel (128 bit) result, you must install one of your two memory modules on DDR3
Appears that on certain nForce2 mobos (which all of these are) Dual-Channel only works with 2 sticks. The original post wasn't very specific regarding the mobo in question.

[climbski]

Suggest some of you READERS try doing some actual BENCHMARKING. Use sisoft sandra bandwidth benchmark and you'll see for your selves that DC dosn't help AMD hardly ANY when OCd.:)

BTW DC only works with 2 or 4 Dimms.

[amd_man2005]

Suggest some of you READERS try doing some actual BENCHMARKING. Use sisoft sandra bandwidth benchmark and you'll see for your selves that DC dosn't help AMD hardly ANY when OCd.:)

I suggest you shut your mouth. AS I have done benchamrks, and also, If there is a VISUAL difference, then why even go to the benchmarks?

[climbski]

Later noobs...LOL Visual... Right. whatever .....people see what they expect. general rule of observation

You obviously havn't done the benches or you'd be agreeing with me. Try lying to someone who dosn't know anything.

Was trying to give the members of this forum some very good advice.... oh well your loss not mine.

BTW game loading times are 90% dependant on Hard drive performance. Which changes all the time for each program due to position and cache status.

[amd_man2005]

Trust me m8, I can tell when Im not in dual channel, AND YES i have done the friggin benches, So i dont need you tellin me I havent. Soo, If you still wanna go calling me and the other members wrong, go ahead, but you'll end up like Soulburner eventually.
:cheers:

[climbski]

What FSB are you running?

BTW I'll end up where I'm at running nearly 2.6ghz 3300 mhz/sec mem bandwidth on single channel 24/7 on and XP1700.:laugh: Didn't get there not knowing what I'm doing.:2cents:

[Wiggo]

The funny thing is that you and Soulburner sound like ya's prefer to use single channel then save a few extra bucks and use VIA chipsets so ya's can go for it while the rest of us just stick with usin' dual channel on our nForce2's.

[climbski]

Sure My VIA chipset NF7-S ver2 gets 233fsb at 2-2-2-6 timings. Why would I want an N2 chipset...Wait a sec dammit I do have an N2 chipset. DOH!

What's your bandwidth wiggy?

I only care about SPeeD my friends and whatever gets me some. DC dosn't on AMD.

POST your Sandra bandwith numbers or stop argueing simple solution eh?

[Wiggo]

Seems that ya have a prob knowin' what ya have anyway. :rofl:

[minibubba]

climbski, you're clearly not going to listen to reason and you're obviously got going to listen to people that actually use the nforce2 dual channel setup and know that it improves performance. So why don't you just drop all this pointless ranting since it's not helping anyone. :no:

[Wiggo]

Well said minibubba.

I think that another noimination is in the air. :laugh:

[climbski]

It'll be an HONOR. I accept ANY future nominations. Thankyou.

Post your bandwidth and shut me up then.:)

If you get more than 6% better BW on DC same settings (postem and I'll duplicate setting in SC to test) I'll eat my words. then we will know once and for all. rigt now all we are doing is steadfastly disagreeing without any data posting.:thumb:

[Wiggo]

The thing is that the 2nd channel does have benefits for performance (and not just in a synthetic bench test) but as ya don't plan on listenin' why bother but I'll take the 5% any day and run with it. BTW I couldn't be bothered to run the tests again as I've better things to do with my time and my PC's but if ya have a look around ya'll find that screenies of most of my setups are posted around these forums so do something useful and search them out for ya self if ya that desparate (I ain't). As stated b4 ya not goin' to change the opinions of those here with those boards so why bother? See ya. :smokin:

[Soulburner]

The dual channel does make a drastic and noticeable difference with the Athlon XP processors.

http://www.ocaddiction.com/forums/showthread.php?s=&postid=147651#post147651

Pretty drastic and noticeable difference huh.....*****es.

[Sovereignty]

Real world gains probably aren't as high as 5%. More like 2.6-2.7%. Some of you guys are saying there's a noticeable difference and that you can easily tell when you're running dual channel as opposed to single? That means you can easily spot the difference between 60fps and 61.5fps.

The link soul posted as to a write up I did on it because I wanted to know what the exact gains were too. My system setup was on an NF7-S with 2x256 OCZ EL PC3500. BTW, neither soul nor climbski have via chipsets. I have no idea why someone would even bring that up.

You can't compare across different chipsets, saying that the NF2 outperforms the Via KT400/600 because it's dual-channel. Yes, it obviously outperforms it, but dual channel isn't the only difference between the two. If you want to compare them it has to be on the same exact board with all settings identical except for the single/dual channel change, like I did.

Memory bandwidth gave a 2.6% increase, Pifast gave a 5% increase, and 3dmark2k1 gave a 2.7% increase when running in dual channel as opposed to single channel. Is there a difference? Yes. Is it drastic or even noticeable in anything other than benchmarks? Not really.