GSkill 1gig LE kit issue on MSI Neo2 with A64 3200+

[drftkng]

Okay, I have this 1Gig GSkill LE kit on my MSI Neo2. I also have a 3200+ winnie that I know will do 2.85Ghz on water. I have done a few iterations of 32meg SuperPi with it at 2.85Ghz on my current setup. This chip has also done 2.85 on a completely different setup completely stable. The only problem is that it isn't completely stable. I have been testing the ram with Memtest86 for the last two days trying to figure out what is going on.

I have basically come down to this, and I am not sure what I should do. As everyone knows the LE kit is supposed to do DDR550 at 2.5-3-3-7. Well I can only get the ram to do 275FSB at 2.5-4-4-8, and it will not go any higher no matter what I try. This is my current and only stable settings at 275FSB

multiplier 10x
1T
2.5-4-4-8
HTT 3
ram volts 2.6
current bios 1.41mod
275Mhz
the chips are 437 version number

The only way I can get it to go higher stable is to set it to 2T, but that kills my memory speed.

I have not messed with ram dividers yet as I would like to not have to. It seems like a lot of people on here have had their LE kits go a lot higher that 275. Do I have a valid reason to want to try and RMA this set since it technically wont' do as advertised. I have posted on GSkills forum and they are very helpful, but I am just not sure what to do.

I hope I got everything, if you need more info please let me know.

jay

[Sup3rman]

I would RMA

[Tirlista]

I have a simliar problem with my LEs. After playing around for a few days I could run 275*9 with no problems 1T. Reading on GSkill site, you'll see the RAM is tested @ 275x9 2.5,3,3,7.

Jack from GSKill reccommends you run 2.5,4,4,8 if you're board isn't a great overclocker.

I've found I can run 2.5,4,3,7 with no dramas - but I can't get my CPU/board combo over ~2600Mhz. I have another Neo2, so I will try that one shortly.

If your CPU had done 2.85 on another setup, but can't do it now, I'd say it's your board holding you back.

[drftkng]

I have had it at 2.85 on this current board, just not super stable. I have another Neo2 on one of my other systems, I am gonna give the ram a shot in that and see what happens.

[glock19owner]

It took about a week or so for my Skills to settle in...see sig for 100% stable, 24/7 timings now...and can bench at 5-3-3-2.5/1T at 275...but not 100% stable...

[drftkng]

Well I just tried it in my other Neo2 board, On this board I was able to get it to do 275 @ 2.5-3-3-7 at 2.7v This was doing about 30 iterations of test #5 in Memtest86 with no errors. I have been running this ram at various speeds for a bit over a month now. Based on this info do you think I just don't have a strong board in the first system, and that the ram is proably alright, just not exceptional for anything over 275.

[glock19owner]

It would seem as though your first board was the weak link...the NEO2's are not doing with the best of luck as of late...lot of people at other forums are having issues running past 260 HT on memory and 2.4Ghz on CPU...

[Zeus]

If the RAM runs memtest errorfree at the rated speeds it's just your other board that isn't willing to coöparate.

Wait for the DFI NF4 to have the bugs ironed out and go with that i'd say.

[Tirlista]

Wait for the DFI NF4 to have the bugs ironed out and go with that i'd say.

That sounds like a very good idea Zeus!

[bito]

I agree with the above....my 1gb 'LE' are running 300HTT at CL2.5/7/4/3 1T @ 2.7v on the DFI NF4 and many others have the same experience.....my sticks are also 0437's. Prime95/memtest stable

[drftkng]

Thanks for the help guys, cause this was driving me If the newer boards are having problems with high oc's that would make sense cause this board came with bios version 1.41 I believe. I previously had the ram in a very early board, but had a bit of an accident with it and it had to be replaced I didn't want to take that board much above 2.65Ghz as I was on air at the time, so I never really tried to push the ram on that one.

jay

[bacon]

iv had 2 sets of LE that mysteriously both topped out at 270 2.5-3-3-7, not 1mh zhigher even if i dropped the timings on my msi neo2, changed the POS to a dfi pcie and what a suprise i hit 305 2.5-4-4-7 with same voltage etc

[ezrael80]

hey guy's
i have also neo2 + Gig LE + 3000+ on water. in my setup (bios 1.41Mod) i can run 1:1 275@2.5-3-3-7 in memtest, but in windows i got massive stability issues. when i put 2.5-4-3-7 it's better, i think neo2 isn't great overcloker after all...

[austinbmxnig]

this is one of the reasons im waiting for the dfi nf3 939 agp board

[glock19owner]

this is one of the reasons im waiting for the dfi nf3 939 agp board

Thought that got axed a few weeks ago...

[robocopy]

My 2x512MB Gskill LE with 440 chips does 9x297 1:1 2.5-3-3-7-1T DC SuperPi-stable.
Try bios v1.36mod. This bios is the best for my gskill.

[ezrael80]

Turez what is difference between bios v1.38Mod and others on your config ?
you can clock higher your Gskill on this bios ?

[t-max]

I have a msi neo 2 and a 1gb g.skill LE kit aswell.
With my winchester 3500 + i was able to hit 284 mhz 2.5,4,3,7 rock stable 1T
then i recived my prommy and changede my winnie to a newcastle becouse of subzero probs. here comes the funny part now i can only hit 275 mhz maks 2.5,4,3,7 .. crap stable with memtest 86.. in windooze all i can do is 266 x 11 = 2926 mhz stable but 275 x 10 is totaly unstable ????.
all this makes me wonder maby its just the ondie memory controller at the cpu that can be crappy ??
otherwise i cant see why a change of cpu´s can couse such big difference in memmory preformance

[Zeus]

I have a msi neo 2 and a 1gb g.skill LE kit aswell.
With my winchester 3500 + i was able to hit 284 mhz 2.5,4,3,7 rock stable 1T
then i recived my prommy and changede my winnie to a newcastle becouse of subzero probs. here comes the funny part now i can only hit 275 mhz maks 2.5,4,3,7 .. crap stable with memtest 86.. in windooze all i can do is 266 x 11 = 2926 mhz stable but 275 x 10 is totaly unstable ????.
all this makes me wonder maby its just the ondie memory controller at the cpu that can be crappy ??
otherwise i cant see why a change of cpu´s can couse such big difference in memmory preformance

It's your memcontroller, no doubt.

[austinbmxnig]

Thought that got axed a few weeks ago...

nah nah