OK. G.Skill 4400LEs came in.

EclipseJP · Feb 8, 2005

My initial results are not great. I can run 2.5 3 3 7 at 2t but it fails every so often. I am thinking that they may need to be burned in. I am not sure though. This is on a Winchester 3000+ 0449.

killerkid · Feb 8, 2005

err correct me if im wrong, which i very well may be, but i thought you were only supposed to run this stuff at 1t??

EclipseJP · Feb 8, 2005

Im not sure. Oh btw they are 0437 chips.

EclipseJP · Feb 8, 2005

Update: Ok these sticks freeze even the bios at 275 2.5-3-3-7.

CpuZ identifies these as follows:

Product Number CL3-4-4-8-Gskill

Manufactured Date Week 02, Year 04

SPD Tables 200fsb 2.5-4-4-8 277fsb 3.0-6-6-11

WTF is that. These cant be LEs.

austinbmxnig · Feb 8, 2005

that guy from g.skill that's always on xtremesystems might be able to help you

EclipseJP · Feb 8, 2005

Signed up on Xtremesystems forum. Now just waiting till I can post. Still no luck with the memory.

MoPMatrix · Feb 8, 2005

EclipseJP said:
Update: Ok these sticks freeze even the bios at 275 2.5-3-3-7.

CpuZ identifies these as follows:

Product Number CL3-4-4-8-Gskill

Manufactured Date Week 02, Year 04

SPD Tables 200fsb 2.5-4-4-8 277fsb 3.0-6-6-11

WTF is that. These cant be LEs.

This is normal, CPUz shows all of the exact stuff for my LE
But mine do what they are rated for 200 2-2-2-5 1t and highest I've tested is 271 2.5-3-3-7 1t,, all at 2.6v. With no burn in. According to the g.skill site these are rated to run from 2.65 to 2.85v.
Ether ur chips are defective or u have some setting messed up, I never ran my chips in 2t though, did u attempt 1t?

rseven · Feb 8, 2005

What a manufacturer rates their ram and the SPD's are two different things. SPD is slower as it's a nice safe setting for the computer novice to set things at so they will have stability and zero problems.

EclipseJP · Feb 8, 2005

Ya ran at 1t settings with every combination of voltages. It will even freeze in the bios when I have it at 275 2.5-3-3-7 1t with 2.8volts. Runs fine with 3-4-4-8 on everything at 275. But when I change the TRC in memtest to 3 it will just freeze up the computer.

rseven · Feb 8, 2005

Eclipse, dheck this out.
http://www.madshrimps.be/?action=getarticle&number=5&artpage=953&articID=244

glock19owner · Feb 8, 2005

EclipseJP said:
Ya ran at 1t settings with every combination of voltages. It will even freeze in the bios when I have it at 275 2.5-3-3-7 1t with 2.8volts. Runs fine with 3-4-4-8 on everything at 275. But when I change the TRC in memtest to 3 it will just freeze up the computer.

Most G. Skills seem to need to settle in for a week or so...

After maybe 2 weeks I was able to run at 275 5-3-3-2.5/1T/1:1 at 2.85v for benching and 5-3-4-2.5 275 100% stable...

Clevor · Feb 8, 2005

EclipseJP said:
My initial results are not great. I can run 2.5 3 3 7 at 2t but it fails every so often. I am thinking that they may need to be burned in. I am not sure though. This is on a Winchester 3000+ 0449.

In my experience, the ram don't have much headroom. 2x512 TCCD peak at 277 or so at 2.5-3-3-7 and 285 or so at 2.5-4-4-7, 1T of course. What does GSkill say: one stick in ten do this?

The 9NDA3+ is a great board, but ram performance varies greatly based on the slots. Try slots 1/2 and then 3/4, then switch positions of the sticks (label them A and B).

On some Winnies, TCCD max out at 269 or so at 2.5-3-3-7, based on the mem controller. The diff between CPUs is something like 3-5 mhz; it's not like the ram will do 269 with one Winnie and 295 on another.

glock19owner · Feb 8, 2005

Clevor said:
In my experience, the ram don't have much headroom. 2x512 TCCD peak at 277 or so at 2.5-3-3-7 and 285 or so at 2.5-4-4-7, 1T of course. What does GSkill say: one stick in ten do this?

They dont warrant LE's rated 7-3-3-2.5 if only 1 out of 10 sticks do this...

1 out of 10 chips make it to the LE standards...not 1 out of 10 sticks

Please post a link or a direct quote from a G. Skill rep that stated 1 out of 10 sticks of the LE's do 7-3-3-2.5 please...

killerkid · Feb 8, 2005

are the LA's alot better?

EclipseJP · Feb 8, 2005

OK. Thanks for the article rseven. My cpu is a newer one and I havent heard to much of newer cpus being finicky.

Clevor- I have them in slots 3/4 and I will try slots 1/2. In fact that just occured to me just before I read this post again.

glock- Thanks for all your help. In fact your the one who convinced me to go this route.

Ok. Will try slots 1/2 and see what happens.

Clevor · Feb 8, 2005

glock19owner said:
They dont warrant LE's rated 7-3-3-2.5 if only 1 out of 10 sticks do this...

1 out of 10 chips make it to the LE standards...not 1 out of 10 sticks

Please post a link or a direct quote from a G. Skill rep that stated 1 out of 10 sticks of the LE's do 7-3-3-2.5 please...

Read my post again: what I said is 2x512 TCCD peak at 275-277, and one in ten do this (according to GSkill), so these are marked LEs.

Go check the GSkill forums, a couple of guys are also having problems meeting specs. I have not found burning in to help TCCD, and I got 4 gig.

glock19owner · Feb 9, 2005

I go to the forum all the times...even posted G. Skills forum on a lot of post on here

...

But you are the only one that has stated that giving the memory to settle in doesnt work...

Do a search and you will see where a lot of members had issues with the G. Skill's at first...but after a week or two...they are running beyond the rated speeds...like my LC's...

EclipseJP said:
glock- Thanks for all your help. In fact your the one who convinced me to go this route.

no problem...If you need any help...you can AOHELL IM me

Clevor · Feb 9, 2005

Well I have my replacement LE sticks on the way from Newegg: it's gonna be show me time.

I'm getting picky now with TCCD and I'm looking for ones that respond to voltage, so you don't have to rely on luck of the draw. My best set of TCCD are the later PQI I got after Newegg sold out of the first batch. This 2x512 pair did 290, 1:1, 2.5-3-3-6 on an AI7 at 3.05 volts (3920 unbuffered). In comparison, the LEs did 271 and did not budge with volts. The PQI can probably do 295 at 3.1-3.2 volts, but wouldn't bench 3D till 281 or so, so no point in going higher. I have not tested on A64 yet but there is a good chance it will bench 3D on this platform.

I've been flogging my first pair of TCCD and they do nothing after several months of burn in. Same with Vitesta PC600. Both use BP PCB too. If anything, I find Ballistix and XMS3500 burn in a little. My worse pair of 2x512 XMS3500 last did 248 or so, 2-2-2-5 at 3.28 volts, but now do 252 - on A64 anyway.

glock19owner · Feb 9, 2005

On this set...write down which TCCD's you have...the last 3 numbers...and did you happen to of noticed what your other sets were?

the LE's should have the 431's, some have had the newer 440's (DDR600 rated)...just make sure you dont have the 437's...and if you do...find out why...those are the poor performers of the TCCD's...

Clevor · Feb 9, 2005

glock19owner said:
On this set...write down which TCCD's you have...the last 3 numbers...and do you happen to of noticed what your other sets were?

the LE's should have either the 431's or some have had the newer 440's (DDR600 rated)...just make sure you dont have the 437's...and if you do...find out why...those are the poor performers of the TCCD's...

The set I RMAed were the highly regarded 0437s. Not only that, every IC on both sticks were 0437. OPB on Extreme says this is his favorite set, so go figure. Hell, guys were killing for 0437s on Extreme.

I'm not taking the spreaders off my PQI but I think my first pair have 0437 and the second 0440. I'm banking the second set of LEs will be 0440 or later.

I'm pretty sure the PQI TCCD Newegg is selling has pretty late stuff.

OK. G.Skill 4400LEs came in.

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Similar threads