DFI DK X48-T2RSB-Plus + D9 GMH + DICE

[CryptiK]

I'm going to start by saying I have never benched ram with DICE before, and I haven't used this board or DDR2 for a little while, so don't expect anything great, I was just having some fun.

I made some ram pots from paper and fitted them to the heatspreaders using electrical tape. In a pathetic effort to protect the board from condensation, I made up some dodgey insulation from paper towel, and filled the gaps around the modules with it.

Wasted 20 mins trying to remember what divider this board likes and some other settings, found 2:3 gave the best results by far. Vdimm was not lowest required for each speed I just guessed for each.

720 MHz @ 2.58 vdimm



731 MHz @ 2.68 vdimm



739.6 MHz @ 2.78 vdimm



I had to stop here as a considerable amount of ice and condensation was forming on the sticks. 739.6 MHz is nothing special, but its a personal best for me. I hope to have another go at this soon.

Here's some pics of the setup:

[Shahryar_NEO]

congrats bro

increase the vDIMM please !!! 2.78v is for air cooling not dice !

btw , really nice pot !

[18 is # 1]

Very cool!
Crank up your AC and it should minimize the condensation.

[Mean Machine]

2.8V isn't much for sub-zero DDR2. Go 3V at least.

[DonNiger]

Yeh, keep it going, my D9 set tied after 3,2v@aircooling.

[BeepBeep2]

Thats awsome, good to see DDR2 being pushed again.

[radaja]

great results CryptiK

[ViViD]

Yeh, keep it going, my D9 set tied after 3,2v@aircooling.

3.2 Air Cooling?

[CryptiK]

Thanks guys.

Pushed a bit harder this morning. Managed to get 751.7 MHz

741.1 MHz @ 2.95 vdimm



751.7 MHz @ 3.08 vdimm



Memtest @ 750 MHz (DDR2-1500) @ 3.08v. No where near stable but looks good to see those numbers. The sticks went from sub zero to positive temps in about 10 seconds so I switched off

[Nanok]

Verry nice ram oc CryptiK

[Mafio]

good work mate

[CryptiK]

Verry nice ram oc CryptiK

good work mate

Thanks guys.

congrats bro

increase the vDIMM please !!! 2.78v is for air cooling not dice !

btw , really nice pot !

Very cool!
Crank up your AC and it should minimize the condensation.

Good point about the aircon, will have to try that next time.

Shahryar, how much vdimm have you used on DICE without killing sticks?

[quake6]

I was using 3v on AIR (ambient was about 6°C) for about 3h without killing... That was on Crucial value (D9 GMH) and with Cellshock (D9 GKX), so you have space to raise vdimm

[Shahryar_NEO]

Thanks guys.
Shahryar, how much vdimm have you used on DICE without killing sticks?

3.1-3.2v on a good air cooling and 3.3v on DICE ( maximum available voltage ) .

I didn't have any problem until now with these voltages ( of course not stable but you could get validation ) .

Btw , if your modules got problem put them in the freezer for a few hours , it'll be ok again ! ( I tested more than 10 times and it worked ! )

try to put DICE directly in front of chips on the Heat Spreader .

[quake6]

And also some chips don't like high vdimm, so test it and you'll know it

[CryptiK]

I was using 3v on AIR (ambient was about 6°C) for about 3h without killing... That was on Crucial value (D9 GMH) and with Cellshock (D9 GKX), so you have space to raise vdimm

3.1-3.2v on a good air cooling and 3.3v on DICE ( maximum available voltage ) .

I didn't have any problem until now with these voltages ( of course not stable but you could get validation ) .

Btw , if your modules got problem put them in the freezer for a few hours , it'll be ok again ! ( I tested more than 10 times and it worked ! )

try to put DICE directly in front of chips on the Heat Spreader .

And also some chips don't like high vdimm, so test it and you'll know it

OK thanks for the info. Good to know I should have more headroom. Most I've used on air was 2.78v while doing some SPI32M runs, and the sticks were still fine after hours of use. I've never killed any ram so I haven't needed to freeze it, but I hope it works for me if it comes to that.

I do also definitely have to make better pots so the DICE sits in front of the IC's, it was only touching the top edge. That should help get some more MHz out of them.

[k0rnh0li0]

nice job :] i'd be scared if it started to freez up the ram slots