Dual channel worth it?, Casualty one ram stick Options

[B82R3S]

Hey guys

For my ITX system i got a P8Z77-I board, and i had 2x8gb ram sticks, i now have one because of what happened when i tried to get the heatspreader off, (didnt use enough heat)

So i now have 1x G.Skill Ripjaws X F3-1600C9S-8GXM 8GB

Will i see a huge benefit in performance from two sticks?

I dont do much stuff with VM's atm, just gaming pretty much... so dont need any more GB of ram, but if dual channel offers a sizeable increase in performance ill buy another stick and have another attempt at getting the heatspreader off.

This is a pic why i needed to remove the heatspreader from one stick (dead stick moments before bin)



This post has been edited by B82R3S: Jun 29 2012, 02:01 AM

[tunksy]

yes dual channel will give you more performance.... go buy some more ram.

[B82R3S]

is it proven/disproven though? or to what extent? because its fine to say yes its better go buy more ram but if the benifit is neglible then i can save the money for other stuff, no offence

[GhostFaceKilla]

There is a small performace boost with Dual Channel. But in gaming it is not really that pronounced.

Here are some old comparisons:
http://www.tomshardware.com/reviews/PARALL...NG,1705-11.html


http://www.youtube.com/watch?v=rgLCEMW-kfY



Lots of stuff on the web.

My advice - if your not running a performance machine with apps that require maximum processing performace then dont bother. You wont see any real difference in gaming.

This post has been edited by GhostFaceKilla: Jun 29 2012, 01:03 PM

[B82R3S]

There is a small performace boost with Dual Channel. But in gaming it is not really that pronounced.

Here are some old comparisons:
http://www.tomshardware.com/reviews/PARALL...NG,1705-11.html


http://www.youtube.com/watch?v=rgLCEMW-kfY



Lots of stuff on the web.

My advice - if your not running a performance machine with apps that require maximum processing performace then dont bother. You wont see any real difference in gaming.

yeah there we go just wanted someone to confirm my thoughts. Thanks,

[Dasa]

benchmarking arma 2 now to see what the difference is on a overclocked 2600k since there is no modern tests

edit
arma 2 does have a fair variance in tests during the same run so this isnt perfectly accurate but the difference seems to be sufficiently large and consistent in these few tests

2600k 1600 9-9-9-24-2t
e1\e8\e2 are the benchmark tests run in arma 2 scenarios

4.3ghz dual chanel ht off
e1 Avg: 42.694 - Min: 29 - Max: 87
e8 Avg: 64.700 - Min: 50 - Max: 91
e2 Avg: 16.659 - Min: 11 - Max: 21

4.3ghz single chanel ht off
e1 Avg: 41.541 - Min: 27 - Max: 88
e8 Avg: 61.071 - Min: 39 - Max: 91
e2 Avg: 16.099 - Min: 11 - Max: 21
~3-6% difference on average fps

4.8ghz dual chanel ht on
e1 Avg: 47.197 - Min: 34 - Max: 92
e8 Avg: 69.088 - Min: 49 - Max: 103
e2 Avg: 19.315 - Min: 14 - Max: 23

4.8ghz single chanel ht on
e1 Avg: 42.887 - Min: 28 - Max: 91
38 Avg: 64.102 - Min: 43 - Max: 99
e2 Avg: 17.756 - Min: 12 - Max: 22
~10% difference on average fps and a fair bit more on the minimum

so not surprisingly the further you overclock the more starved for memory bandwidth the cpu becomes
this is also evident in memory speed tests comparing 1333 vs 1600 and higher


back in the days of ddr1 dual channel made ~7% difference its interesting to see that hasnt changed much

This post has been edited by Dasa: Jul 1 2012, 02:10 AM

[B82R3S]

interdasting, performance vs cost isnt that bad, might consider it when i feel lucky enough to try taking the heatspreaders off again

[xyzzy frobozz]

interdasting, performance vs cost isnt that bad, might consider it when i feel lucky enough to try taking the heatspreaders off again

Or you could just go here:

http://www.downloadmoreram.com/

;-)