2x1GB OCZ PC4000EB Platinum + DFI NF4 UT SLI-DR

[eva2000]

Pics: http://www.fileshosts.com/memory/ocz/2GB_PC4000EB/



Still using 623-3 bios

FX-57 0516 WPMW Water cooled
DFI NF4 UT SLI-DR R.AB0 623-3 Official bios
1x BFG 7800GTX 81.82 beta drivers
4x 80GB Hitachi 7K80 SATAII Raid 0
1x 250GB Maxtor Maxline Plus II PATA
Pioneer DVD Burner
Floppy drive
650W Silverstone SLI PSU

WinXP Pro SP2
NF4 6.53WHQL chipset drivers

12x250

10x280

10x282 <-- max i've tested so far 32M stable

Did some 3dmark tests to confirm OCZ's claims that max 3dmark/gaming speeds will be lower than max superpi capable speeds due to most AMD64 mem controllers being stressed by 2x1GB with 3-3-2-8 timings.

Interesting findings

max memtest stable = 283mhz 3-3-2-8 at 2.64v
max superpi 32M stable = 282mhz 3-3-2-8 at 2.64v

but tried running 3dmark2003 at stock BFG 7800GTX 460/1300 speeds and this is what i got:

vcore = 1.4v (1.36v cpuz)

10x280mhz 3-3-2-8 at 2.64v = freezes in GT1 tried several times always freezes in GT1

10x280mhz 3-3-2-8 at 2.69v (2.6v + 0.03v option) = freezes in GT4 nature test

vcore = 1.425v (1.392v cpuz)

10x280mhz 3-3-2-8 at 2.69v (2.6v + 0.03v option) = PASSED but score was lower than usual 17400ish, rerun it again and PASSED again but now score was 13400ish ???

Adding a bit more vcore helped complete 3dmark2003 but score is weird ??

Seems OCZ is correct that 3dmark capable speeds will be lower than max superpi 32m speeds.

 

[eva2000]

Now testing 11x274mhz 3-3-2-8 to see how it goes.

11x274mhz 3-3-2-8 definitely works ran GT1 and GT4 (both most stressing on system) in 3dmark2003 in 40 loops @1600x1200 resolution with no probs

More tweaks to come hopefully

 

[darksparkz]

Whoa, those are some pretty nice speeds with a 2x1GB set, timings aren't too bad either.

 

[RedDragonXXX]

Looking good and that's a prime example why people by FX's because they have insane memory controllers!

I'm assuming those are Samsung UCCC's or the latest batch of CE6's, correct me if I'm wrong.

 

[CCUABIDExORxDIE]

thse are the latest batch of CE-6. i was baffled by it, as i didnt they that CE-6 could do 3-2-3. and the use of the EB name made me think micron. those are some awesome clocks though! keep up the good work.

 

[Zebbo]

eva2000, have you tested out highest 3D stable clocks yet?

 

[eva2000]

11x274mhz 3-3-2-8 at 2.64v is highest i've tested 3dmark stable so far haven't tested higher yet

 

[KAMALINDA!]

cool