OCZ 2GB PC4000 GOLD XTC testing

[perry_78]

This should have been done AGES ago...

These sticks came directly from OCZ, are Samsung UCCC chips, however my CPU mem controller appears to have trouble running over 275mhz 1:1 with a 10x multi without a compromise in stability. I've tried many max async/read pream and drive strengths, but in vain Not sure what weekcode, bigtoe told me that I shouldn't venture into the lands of heatspreader removal unless I like ripped off chips.

So far I have no results at stock speeds, to be honest I never even ran these stock All vdimm values are BIOS set values! Real values are a little off, the sweet spot seems to be exactly 2.8vdimm (2.7v + 0.3v bios set).

ddr540 3-4-4-8 2.6vdimm SUPER PI 32M
ddr540 3-4-4-8 2.6vdimm SANDRA MEMORY BENCHMARK
ddr540 3-4-4-8 2.6vdimm SANDRA MEMORY + CACHE BENCHMARK
ddr540 3-4-4-8 2.6vdimm EVERLAST LATENT
ddr540 3-4-4-8 2.6vdimm EVERLAST READ
ddr540 3-4-4-8 2.6vdimm EVERLAST WRITE


Bumped the vdimm a bit, raised TRAS and lowered multi for stability's sake, this 3700+ is nothing like the old opteron

ddr540 3-4-4-10 2.7vdimm SP2004
ddr540 3-4-4-10 2.7vdimm SUPER PI 32M

Next step was of course ddr550. This time another voltage bump with+0.3v, which I found was the optimal voltage for this particular kit, and TRAS to 12. For some reason, TRAS helped a lot. I'm assuming this is because I screwed some other setting I am not aware of, as never has TRAS had such an impact before from my experience. Any advice will be greatly appreciated!

ddr550 3-4-4-10 2.7 +0.3vdimm SP2004 running
ddr550 3-4-4-10 2.7 +0.3vdimm SP2004 done
ddr550 3-4-4-10 2.7 +0.3vdimm SUPER PI 32M
ddr550 3-4-4-10 2.7 +0.3vdimm SANDRA MEMORY BENCHMARK
ddr550 3-4-4-10 2.7 +0.3vdimm EVERLAST LATENT
ddr550 3-4-4-10 2.7 +0.3vdimm EVERLAST READ
ddr550 3-4-4-10 2.7 +0.3vdimm EVERLAST WRITE

Another TRAS bump. This was easily 3d stable, and I think I can get this to run SP32M as well once I figure this controller out. I need to re-run 3d though, only have a 3-4-4-12 275 screenie.

ddr560 3-4-4-12 2.7vdimm +0.3v SUPER PI 1M
ddr570 3-4-4-12 2.7vdimm +0.3v SUPER PI 1M

addition meanwhile:

ddr550 3-4-4-12 2.7 +0.3vdimm 3DMARK06

-> currently wr in x600xt, first attempt

I hope to get more results at DDR400, DDR500, 2.5-3-3-6 etc. later in the near future! 3dMark may be an option, however most of my gfx's are currently busted

btw, if you want to host these images internally, please do! I hate imageshack.

[edit] going for ddr400 later, CAS 2 is a no go.

[tHe_jOkeR]

so max. stable is?

[perry_78]

Currently undetermined, somewhere between 275-280 I would think. 280 seems probable enough to me.

I also noticed that these sticks clock MUCH better with active cooling than without. Even at 2.5vdimm these were blazing hot

[ozzimark]

nice work perry

just have one thing... what's the max mhz you can do with the tightest timings obtainable at 200mhz?

thanks

[bachus_anonym]

Great job, Perry

Thanks for posting your testing results in XB because it's been getting a bit lonely here, without OCZ tests I'd like to see more brand variety and quality tests in this section. Keep up the good work

[perry_78]

nice work perry

just have one thing... what's the max mhz you can do with the tightest timings obtainable at 200mhz?

thanks

I'll re-run and report, I think it was around ~235 2,5-3-3-6, but dont quote me on that

thanks bachus, it still needs to be expanded, you're reports set the standards

[Opteron96]

Hi.
I`m dual priming at 3-3-3-8@250MHz. 2.7V
Update soon.

[Breeze]

excellent work

[Capt_Caveman]

Thanks for the info. I just recently picked-up the kit but haven't done any oc'ing yet.

[Opteron96]

Her it is.

19H dual prime.
260MHz 3-3-3-8 2.9V.

I was trying 270MHz thats why 2.9V. I did 9H 3-3-3-8 2.8V erlier.

[IMG]<a href="http://xs.to/xs.php?h=xs73&d=06124&f=19t_prime_260_3_3_3_8.PNG" target="_blank"><img src="http://xs73.xs.to/pics/06124/19t_prime_260_3_3_3_8.PNG.xs.jpg" title="Free image hosting powered by xs.to"></a>[/IMG]

[ozzimark]

are those gold sticks, or the platinum EL XTC?

[Opteron96]

Gold

[ozzimark]

wow, i guess i've never really seen UCCC pushed at 3-3-3.. though now that i think about it, i think ocz was putting some 5b F in certain gold sticks. do you have rev 1.1?

[bachus_anonym]

wow, i guess i've never really seen UCCC pushed at 3-3-3.. though now that i think about it, i think ocz was putting some 5b F in certain gold sticks. do you have rev 1.1?

With this type of voltage (2.9v) it's may be quite possible to see what Opteron96 did with his OCZ Gold. I have not given my UCCC more than 2.73v (DMM) for 3-3-3 but @ 2.68v (DMM) it was priming @ 250MHz 5-3-3-3.0 and last time I checked it did 255Mhz in SPi32M. I have them in ATM and will see if 260Mhz is doable - it's not OCZ, though...

BTW, very nice Opteron96 Can you see what they do @ 8-4-4-3.0-1T, please?

[ozzimark]

good to know, thanks bachus. one of these days i'm gonna get myself a set of UCCC to play with, just so i have a better idea of what to say when trying to help people out

[Opteron96]

With this type of voltage (2.9v) it's may be quite possible to see what Opteron96 did with his OCZ Gold. I have not given my UCCC more than 2.73v (DMM) for 3-3-3 but @ 2.68v (DMM) it was priming @ 250MHz 5-3-3-3.0 and last time I checked it did 255Mhz in SPi32M. I have them in ATM and will see if 260Mhz is doable - it's not OCZ, though...

BTW, very nice Opteron96 Can you see what they do @ 8-4-4-3.0-1T, please?

This is where it gets werd. 3-4-3-8 and 3-4-4-8 dont get me higher!

strange

But im still testing.

[bachus_anonym]

This is where it gets werd. 3-4-3-8 and 3-4-4-8 dont get me higher! strange But im still testing.

Hmm... Can you do us a favour and post a CPU-Z SPD Tab and a pic of that kit? That would be very helpful Thanks!

[perry_78]

3-3-3-8 is pretty slick :o Mine need a tras much higher to maintain any stability in Sp2004. I think I can get some time to re-run 3-3-3-8 tomorrow, but I wouldnt get my hopes up

[Delirious]

Sounds exactly like my set, maxs out around 278 2.6v 3-4-4-8.

Currently undetermined, somewhere between 275-280 I would think. 280 seems probable enough to me.

I also noticed that these sticks clock MUCH better with active cooling than without. Even at 2.5vdimm these were blazing hot

[ozzimark]

along with what bachus said, look for that revision 1.1 label i mentioned earlier!
i might be jumping to conclusions too quickly, but no gains over 3-3-3 is very typical of micron dies

[perry_78]

No revision label - PN ends in -k.

Hence I assume I have UCCC sticks, as 3-4-4-12 gives great mhz gains over 3-3-3-8.

[hlonipha]

I've tryed to get my sticks to ddr550 following this examples to no avail, I have an expert board with an opty 170.

Can someone with same board show the BIOS settings not the a64 tweaker settings ?

Thanks

[Opteron96]

Here is 265 3-3-3-10 32Mb phi.
Fighting to get it to manage 24H dual prime


[IMG]<a href="http://xs.to/xs.php?h=xs73&d=06125&f=32MB_pi_265_3_3_3_10-2.9.PNG" target="_blank"><img src="http://xs73.xs.to/pics/06125/32MB_pi_265_3_3_3_10-2.9.PNG.xs.jpg" title="Free image hosting powered by xs.to"></a>[/IMG]


Perry 78: I dident ment to hijack this tread.

[bachus_anonym]

I hope to get more results at DDR400, DDR500, 2.5-3-3-6 etc. later in the near future! 3dMark may be an option, however most of my gfx's are currently busted

It would be cool to see how far your kit goes at 5-3-3-3.0. When you get a chance, If you're not tired testing it yet

Here is 265 3-3-3-10 32Mb phi. Fighting to get it to manage 24H dual prime

http://xs.to/xs.php?h=xs73&d=06125&f...3_3_10-2.9.PNG

Man, that is really good for UCCC (if that's what this set is) What voltage is it at? I tried my UCCC yesterday and it passed SPi32M @ 260MHz 5-3-3-3.0-1T, 2.90v (DMM) but haven't tried higher...
Did you manage to work on getting it higher at 7-3-4-3.0-1T? If it doesn't go there it would be quite odd, considering there are supposed to be UCCC...

[Opteron96]

Hi.

3.0-4-3-8-1T dont give me more Hz. I know its strang

Now I need a break form OC.

The 265MHz was with 3.0V