LP UT 250GB with 512MB TCCD 1T DDR 600 BIOS Setup Example

[ugp]

I shall try this out and see what I get...

[ugp]

I couldn't even get one stick with those settings with my HTT at 275...I think I am stuck with a crappy CPU is what my problem is.

[Catalan]

ugp, avatarelectronics sells 3200+ AR if you want a better cpu. That's where I got mine.

[ugp]

Can't afford it but here is what I got so far from his settings...

275x8.0 1:1 with 2 DIMMs

It was a 41s SuperPi time.

[Catalan]

41s 1M in SuperPi? If so that seems kind of low, I get 36 with my bh-5 @ 240x10, 1:1, CPC ON, 2-2-2-5.

[ugp]

Well my CPU is only at 2.2 and look at my timings as well.

And another thing to add...If I put my RAM in DIMMs 1 & 2 the system wont even post. For some reason it doesn't like DIMM 2. I my RAM in DIMMs 1 & 3. So far 275FSB seems pretty stable and that is about what you are going to get out of Plat Rev 2 with 2 Sticks in. I have seen some people with 285FSB but I am not sure if that was 100% stable or not.

[Walrusbonzo]

For me the max I can do totally stable(Prime95 and MEMTEST86 for 8+ hours) with 2 DIMMs(2x512mb Patriot PC3200+XBL with Samsung TCCD) is about 255~260MHz, this is with timings of 2.5,3,3,7 at 2.7v. More voltage and slacker timings does not help at all. I have to use slots 1 and 3 as 1 and 2 are unstable together, as are 2 and 3 together.

With 1 DIMM in slot 1 things are a little more strange. I'm perfectly stable for 8+ hours in MEMTEST86 at 300MHz 1:1 with everything set as Oskar_WU has stated, but I have EXTREME difficulty getting into Windows, often it locks while booting. More chipset voltage appears to help, while nothing else appears to help get it stable. Further testing reveals the same problem at even 280MHz 1:1. The one time I did get it stable in Windows for more than a minute was at 290MHz, Prime never quit with errors, but instead just hard locked as it normally does while booting.

Think it might be the CPU holding me back, and if it ain't that it must be the mobo, I'd be VERY surprised if it was the RAM.

[ugp]

Yeah with me it is my CPU holding me back. My memory controller isn't as good as what some other people have. I wish I had a better one

[spaceman]

id say, according to jess, 2-1T-2-10-2-8-12-2-2-1-2-4708-1 and then the other settings with all enabled

Thanx dude, I guess I missed that before. Didn't get me any more FSB, but a little more bandwidth, in SANDRA.

[mad mikee]

For me the max I can do totally stable(Prime95 and MEMTEST86 for 8+ hours) with 2 DIMMs(2x512mb Patriot PC3200+XBL with Samsung TCCD) is about 255~260MHz, this is with timings of 2.5,3,3,7 at 2.7v. More voltage and slacker timings does not help at all. I have to use slots 1 and 3 as 1 and 2 are unstable together, as are 2 and 3 together.

Think it might be the CPU holding me back, and if it ain't that it must be the mobo, I'd be VERY surprised if it was the RAM.

When you were running 2 dimms, did most of your memtest errors occur above 512m when you pushed FsB (Thats What I get and I think we are running into the 2dimm problem for AMD chips)

[cargo555]

AMD64 3400+ CG 1mb Clawhammer
XP-90 HeatSink + 92mm Panaflo
DFI 250gb with modded BIOS
2x512mb Gskill PC4400 rated 2.5-3-3-7 TCCD

I too cant OC 2x512mb of the TCCD with 1T too much....245-250mhz 1:1 seems to be max that these dual sticks @ 1T can do without errors in memtest...at 2.5-3-3-10 @ 2.9V

Errors would always appear at 96-97% of Memtest Test5 loops....producing thousands of errors

individually, these sticks run fine, and each stick by itself is able to run 275mhz 2.5-3-3-6 2.7V no memtest errors at all (1T)

but together, OC seems to hit a wall when it comes to 1T


could it be a problem with the on-die mem controller? How much OC headroom do the ClawHammers have @ 1T with 2x512mb?

[Catalan]

Usually with 2x512 you use 2T since they're double sided modules and the board doesn't respond well to it since it's like having 4x256 modules. For maximum OC people usually use 1x512 or 2x256.

[ugp]

That is why you get OCZ Plat Rev 2 People are doing 285+ with 512x2

[cargo555]

well it also has a lot to do with the ability of the CPU's on-die memory controller to handle 285+mhz with 2x512mb

[chunkylover77]

Is OCZ Plat Rev 2 better than 3700eb?

[eva2000]

on average single stick of OCZ PC3200 Plat Rev2 will clock higher than OCZ PC3700EB although a few PC3700EB from earliest batches clock exceptionally high

[Catalan]

Cargo it has been known since back in the AXP days that 2x512 modules normally require 2T to achieve high clocks.

[p4z1f1st]

Oskar Wu, could you tell me, which BIOS you use / recommend?

and could someone here tell me, where i find the modded bioses?

[STEvil]

i have a dumb question: is all tccd this good?????

i've got 2x256 of tccd i bought quite a while ago and its doing 250fsb (max of board) on the ga-k8vt800 next to me...

Not bad for only 2.7v?

gotta hook up a cdrom drive and do some memtest now

[p4z1f1st]

AMD64 3400+ CG 1mb Clawhammer
XP-90 HeatSink + 92mm Panaflo
DFI 250gb with modded BIOS
2x512mb Gskill PC4400 rated 2.5-3-3-7 TCCD

I too cant OC 2x512mb of the TCCD with 1T too much....245-250mhz 1:1 seems to be max that these dual sticks @ 1T can do without errors in memtest...at 2.5-3-3-10 @ 2.9V

Errors would always appear at 96-97% of Memtest Test5 loops....producing thousands of errors

individually, these sticks run fine, and each stick by itself is able to run 275mhz 2.5-3-3-6 2.7V no memtest errors at all (1T)

but together, OC seems to hit a wall when it comes to 1T


could it be a problem with the on-die mem controller? How much OC headroom do the ClawHammers have @ 1T with 2x512mb?

same Prob here....i think, my 2800+ Memcontroller can't handle them....

but i've tested the last hour ONE stick with Oskar's settings and that came out

ABSOLUTLY MEMTEST Test#5 @ loop for 30min STABLE



just set for fun with ClockGen

[Ubermann]

Thats 2T right ?

[JWB]

Finnaly we see over 300fsb with this RAM by an XS member apart from onepagebook Show us your max clock please

[Walrusbonzo]

When you were running 2 dimms, did most of your memtest errors occur above 512m when you pushed FsB (Thats What I get and I think we are running into the 2dimm problem for AMD chips)

Yep, spot on, most, if not all the errors were in the last 512mb, I.E. the second DIMM.

[p4z1f1st]

Thats 2T right ?

nope - 1T (look at A64-Tweaker )

i mean, how shall i get nearly 5000MB/s @ SingleChannel with 2T?

[Mrki]

nope - 1T (look at A64-Tweaker )

i mean, how shall i get nearly 5000MB/s @ SingleChannel with 2T?

with ocz eb's?