Gigabyte X48T DQ6 Living Review

[Ace-a-Rue]

Originally Posted by supershanks

Well has taken a bit longer than normal , but have about bottomed the usual Q6600 9x400 clock I think. Was going to prime it today , but the postman bought an E8200 that i'd ordered to go in the wife's PC. Well I've got to test it for her haven't I Muhahaha So the Q6600 clocks will wait for a bit

So just to make sure it works I'll run a few clocks
If your unfamiliar with E8200 see:-
E8200 Specs

E8200 8x400@4.00B 1600Mhz DDR3

Started off at 8x400 with 4.00B divider giving DDR3 1600Mhz timings are as I'd sussed out for the Q6600 clocks 7-6-5-18 1.85v

Didn't start to good though wprime ran fine , 32M SuperPi bombed

i would not try command rate of 1T...i would go right to 2T...most memory modules are designed around that rate....the only way 1T works well is when the memory controller is right on the cpu die.

[Clunk]

1T where is this 1T you speak of?

[supershanks]

From what I found on the asus P5e3 & so far on here 1T is easy , giving abit extra for little loss of stability. I've also gone from this The Gigabyte GA-X48T-DQ6 - Redefining the High End? - Command Rate (CMD)

The options are Auto, 1N and 2N. A setting of 1N can be used almost exclusively while overclocking, bringing a performance advantage of around 2ns to memory access latency. If memory speeds over DDR3-1800 are desired then a setting of 2N may be required for stability.

managed 4040.23 MHz (505.03 * 8) seems that CPU-Z doesn't get the memory right with 3.33D Divider Ram was actually at 1682Mhz Got about 8 loops into Pi32m then failed. Will pick up with that later on

luck

[Ace-a-Rue]

Originally Posted by supershanks

From what I found on the asus P5e3 & so far on here 1T is easy , giving abit extra for little loss of stability. I've also gone from this The Gigabyte GA-X48T-DQ6 - Redefining the High End? - Command Rate (CMD)

luck

i do not agree with his statement...the only reason he make's that statement is because he gets the best hand picked memory...most of us 'johns' do not get the opportunity to buy hand picked memory...i tried 1N and it would not even post!

[Clunk]

Ah yes, no hand pickeyness here, just good old fashioned retaileyness

NB - If someone wants to send us some hand pickeyness, that will be splendid

[Ace-a-Rue]

YEAH!...then, send some my way

[supershanks]

NB - If someone wants to send us some hand pickeyness, that will be splendid

&

YEAH!...then, send some my way

Nowt wrong with that

Wonder if it's perrhaps down to a limitation on the 2x2gb kit ??
His ram used in the Anantech article :-

OCZ DDR3 PC3-14400 (DDR-1800) Platinum Edition
2x1GB XP, 4x1GB Vista 64-bit

but probably Hand Plicked

[Ace-a-Rue]

definitely 2 x 1gb is better than 2 x 2gb...if i wanted to jam the volts up its butt, i probably could use 1N...i really do not want to go over 2.0.

[supershanks]

Yep I've kept voltage within spec so far on these , my top end is 1.90v

[supershanks]

Initially this wasn't stable when running any load in XP SuperPI 32M for example froze after 4 or so loops. In order to see if it was ram or cpuu causing issues , so dropped the divider and tried again.

7.5x510@1360Mhz DDR3
Click to Enlarge


So I dropped the command rate from 1N to 2N & went back to 3.33D , 1700Mhz DDr3.
Seems like that anandtech claim

The options are Auto, 1N and 2N. A setting of 1N can be used almost exclusively while overclocking, bringing a performance advantage of around 2ns to memory access latency. If memory speeds over DDR3-1800 are desired then a setting of 2N may be required for stability.

from earlier might be based on hand picked ram I guess it puts me firmly with

Originally Posted by Ace's

.most of us 'johns' do not get the opportunity to buy hand picked memory...

7.5x510@1700Mhz DDR3
CPU-Z
Click to Enlarge