OPB's perspecive about The "real truth" of 5BF IC

hello everyone; I have been working on how to set the Micron 5bf IC;
basically, I just figure out someone said is absolutely NOT ture!
The reason I say this is because I have 100 sets of team group 2gb cronus
micron 5bf ic here. I have been testing about 30 sets, 19 of them all passed not only 250mhz 3-3-3-8 but DDR540 for quite nice setting; if you are talking about 250+, then most of them passed 265 32mb spi on 2.7v.
Maybe you guys doubt why previously some of friends bought TG's 5bf and got bad results;but forgive me say this.----> You guys have no IDEA or just think5bf is kinda like 5Bd.

Like I said, and I don't wanna argue if all my rams are all cherry pick. From now on, I will do the serious check before all this rams go to retail. Let's go check the truth and my results on these two days (over 25 sets)
here are the points you guys gotta be aware of:

1.If you are Opteron user, please DO NOT use divider on running 5bf IC. Why, because OPty is well-known
lack of divider of good results.

2. Keep the settings I post here, you don't have to try any loosing timings, 5bf is a bit weird IC;
it needs a special treaments;), I tried many different settings, nothing help on 5bf, you will see.
for ex: the MAL/rp on the auto setting under bios would be 8/6 or even 9/7 after you get into windows
but I tell you what, 8/6 or 9/7 won't let me pass instead of 7/5, weird huh?:D

3.Don't try any weird TREF, only target on 3:
3120: 200mhz, 15.6us----->set this under bios for mainly using purpose.
3684: 100, 1.95----> if you have problem for 3120, then try this one
200, 3.9---> if you really no luck on both above, try this one;), most likely you won't until you try over 275 fsb.

4.Watch your PSU rails, especially on 3.3v; don't tell me you don't believe this,
PSU is the only concern when I bench.
** Be careful for MBM5's 5v reading, which is not correct for reading some big power unit 5v rail but 5vsb;
as I know, my zippy 950w and PC P&C 1kw are like that.
let's see some examples: these result are NOT special pick, they are randomly select from the stock I promise.;)

ddr530 16m testing:

http://img95.imageshack.us/img95/34/...30pi16m8fm.jpg

how about ddr540 32m pass

http://img93.imageshack.us/img93/339...5bf32mp6jm.jpg

well just read this and be patient for playing this, you would be fine.
get used into it. you will understand its patent such as some fixed settings that I figured out.
I hope this will help those who used to be not happy with it.

update, Running lots of Opteron bench, all forgot to grab screen shots,
here you go, nothing different settings from the info above.:D

http://img409.imageshack.us/img409/8...46tg32m1qg.jpg

3d05 single 7900gtx also no problem on exactly settings:

http://img149.imageshack.us/img149/5...0op3d051lo.jpg

And I seriously apologize about the bad experience you guys had before.

OPB

Nice little guide here OPB and thanks for putting in soo much time in order to give the knowledge to everyone whom has been struggling.

Tim

i just got a random kit from TankGuys to play with for a few days. 270mhz is as easy as baking cookies, even with 4x1gb :toast: (actually, 4x1gb clocks a touch better, but i'll get into that when it's time :D)


some things to watch out for though, sometimes tighter timings are better. it seems to me that as the voltage goes up, trrd and twtr MUST go down to maintain stability

also, if you choose to use yellow slots.. for me at least, drive strength must be raised to maintain stability

Very nice Kevin this will help alot of people here. Good Job!!

OPB, since you have over 100 sets, feel free to send me a set, preferrably a good clocking one. :D

Thank you guys, one thing I 'd like to clearify is that:
Yes, the truth is that 5bf might be tough for tuning; however, it doesn't mean
ALL the 5bf will need to be sentenced to death.:D Especially if you place it under the cpu with good memory controll and decent PSU;)

@tsuehpsyde
basically, those pairs can do over 265+, all will be sent to tankguy for sure.
I will NOT allow any sinigle pair fail on that. BUT, just don't blame on the ram if you can't make it,
narrow the prob, find the reason failed could be occured on your components as well.

Quote:

---

Originally Posted by Onepagebook

Thank you guys, one thing I 'd like to clearifyis that:
Yes, the truth is that 5bf might be tough for tuning; however, it doesn't mean
ALL the 5bf will need to be sentenced to death.:D Especially if you place it under the cpu with good memory controll and decent PSU;)

---

it is indeed exceptionally picky, though i think you just made it a lot easier to tweak for many people :toast:
to me, the key with 5b F is to try everything :D

i'm gonna go off on a bit of a side note here, and point out that i've seen nothing good from asus with F dies. just something to keep in mind ;)

ozzimark are you running 2T with 4 sticks?

OPB I dont think you will have any probs with that psu you have:)

yeah.

i'm having a problem getting a screenshot though, i think 270mhz is a bit over my memory controller's limit with 4 sticks in... made it to loop 23 in 32m, and it locked. was priming for about 20min, started opening cpu-z screens.. and it locked. oh well. guess i'll do lower mhz with less voltage and tighter timings to make it equally cool :D
a weird thing though.. i can easily get prime to not fail in blend mode at 280mhz. i'll have to sorta demonstrate later ;)

As always opb comes thru where he should, nice work and effort put forth to bring some light back on the Cronus.

expert is supposedly able to do 1t with 4 sticks in correct?
Have you tried any of this OZ, id like to see some results at any speed just to know that it actually does work and its not all some smoke up da but :)

it can, though at really low mhz.. like 201mhz :p:
also, what is it that you want to see, some results with 4x1gb? ;)

yah 4X1g 1T is all I wanna see :)
I heard expert can do it but I thought this was due to a mem controller issue with amd, now I hear its mobo based, and expert addresses it.
Makes me sad I sold all my bh-5

drive strength is the thing that controls 1t timing, i believe. With high drive strength one can do 1t.

nice guide
I used the exact same tweak with my new ballistix and it's stable windows memtest @ 295fsb 3-3-3-8 2.7v :)

Quote:

---

Originally Posted by ozzimark

it can, though at really low mhz.. like 201mhz :p:
also, what is it that you want to see, some results with 4x1gb? ;)

---

i do! me sir!:clap:

hey OPB, i appreciate the time & effort you've shown here. but i think there are two varieties of concern with these modules. this guide addresses the first variety of concerns, which was that a lot of people were unable to get them stable in the 500-530 range. hopefully this will help, but i notice the modules you're testing in the screenies are week 10, whereas a lot of people were having problems with week 06. did your results hold for that batch, as well?

the second concern is that 540 isn't nearly as spectacular as the review modules sent out at the end of last year which were hitting 560-600. my modules will go up to 550 (maybe), but now i'm running them on a divider at 520 because they won't go 1:1 with my cpu (and for the record, they are very stable despite the use of a divider). now obviously 520 is not very exceptional, and i could have bought much cheaper memory for comparable results.

also, ozzimark is right about asus boards and this memory. you need to make clear to people (and ask your distributors to do the same) that they will need to get a different motherboard if they don't want to run 2T.

Quote:

---

Originally Posted by kovboi

hey OPB, i appreciate the time & effort you've shown here. but i think there are two varieties of concern with these modules. this guide addresses the first variety of concerns, which was that a lot of people were unable to get them stable in the 500-530 range. hopefully this will help, but i notice the modules you're testing in the screenies are week 10, whereas a lot of people were having problems with week 06. did your results hold for that batch, as well?

the second concern is that 540 isn't nearly as spectacular as the review modules sent out at the end of last year which were hitting 560-600. my modules will go up to 550 (maybe), but now i'm running them on a divider at 520 because they won't go 1:1 with my cpu (and for the record, they are very stable despite the use of a divider). now obviously 520 is not very exceptional, and i could have bought much cheaper memory for comparable results.

also, ozzimark is right about asus boards and this memory. you need to make clear to people (and ask your distributors to do the same) that they will need to get a different motherboard if they don't want to run 2T.

---

Good points:

Answer are:
1.that date code you see merely stamped as Team Group package dated not Micron manufactured date. I can just tell in public;) all Team Group's IC date code is 0538 no more.


2.I just got Asus An832SLI DX, I think this must be the one some other peeps wanna try or already tried; I can honest tell you, I can get it work @265~270 1t (the same pair I passed spi32m on DFI both ultra-D and Venus)BUT only doing nothing ; all you need to do is change the memory freq to limit as well as first part of memory parameters, that is, the primary timings.
others are all set as auto.
After you boot into windows, use a64tweaker to change the timings and test them one by one. Well, I don't sound Asus is not the board to run anything , but since Asus's first K7 board until know, the only goodness of AMD was Socket 940 Via board--->k8v delux? Asus is still only good for intel.

so ya, here are the picI can show you, all the micron 5bf teamgroup 2gb cronus in US are here;)
I test them all by from 1m, 8m, 16m, 32m; if it pass, then I sign for that on the sticker. Simply I just don't want to disappoint all the friends who support me.

http://img306.imageshack.us/img306/6994/5bfs5pw.jpg

http://img87.imageshack.us/img87/7760/5bfpass1qc.jpg

Quote:

---

Originally Posted by Onepagebook

2.I just got Asus An832SLI DX, I think this must be the one some other peeps wanna try or already tried; I can honest tell you, I can get it work @265~270 1t (the same pair I passed spi32m on DFI both ultra-D and Venus)BUT only doing nothing ; all you need to do is change the memory freq to limit as well as first part of memory parameters, that is, the primary timings.
others are all set as auto.
After you boot into windows, use a64tweaker to change the timings and test them one by one. Well, I don't sound Asus is not the board to run anything , but since Asus's first K7 board until know, the only goodness of AMD was Socket 940 Via board--->k8v delux? Asus is still only good for intel.

so ya, here are the picI can show you, all the micron 5bf teamgroup 2gb cronus in US are here;)
I test them all by from 1m, 8m, 16m, 32m; if it pass, then I sign for that on the sticker. Simply I just don't want to disappoint all the friends who support me.

---

then i do wonder why almost everyone i've seen with an asus board and 5b F sticks was having serious problems. :shrug:

however, that must be exceptionally time consuming... testing all the sets in sPI.. i'm assuming at 250mhz? that shows some serious dedication to quality :D :toast

ps- if you don't mind, i'm gonna start up a thread with my own results with the 0609M dated sticks i'm testing now.. should have some useful info by tonight :cool:

Quote:

---

Originally Posted by ozzimark

then i do wonder why almost everyone i've seen with an asus board and 5b F sticks was having serious problems. :shrug:

however, that must be exceptionally time consuming... testing all the sets in sPI.. i'm assuming at 250mhz? that shows some serious dedication to quality :D :toast

ps- if you don't mind, i'm gonna start up a thread with my own results with the 0609M dated sticks i'm testing now.. should have some useful info by tonight :cool:

---

ya, from 250mhz then 260mhz if both all pass, then directly jump up to 270mhz; if failed go down to 265mhz---> that's my first 5 pairs test when I struggle consecutively about 24 hours non-sleep. After I found out the patern, I just simply plug pair in, boot up and test it. So if someone got those passed pair and told me that is not even boot at 250 on dfi, I will invite him bring that pair to Columbus, Ohio, and I can test in front of him.;)
sure I don't mind you start that.At least you are the user who really get in study of this; shame on the one who just BS and never tried any effort.

thanks kev :D
though these sticks seem weird, even by 5b F standards. i'm gonna do a bit of work at the lower mhz and tighter timings for those guys who don't wanna run high mhz for whatever reason :toast:

and see something wrong with this? :lol2:

example on Opteron, 1:1, ddr540 rocks stable.
Every single setting is just like what I said. Nothing new
the vdimm showing on the mbm5 actually is what I set 2.7 under Ultra-D BIOS,
Now in the middle of process on super pi 32m running!

http://img342.imageshack.us/img342/5004/op132m016pe.jpg

http://img86.imageshack.us/img86/2504/op132m029ke.jpg


************************************************** **********************************

Updated: see anything different from the setting I provided? Nope!!:D

http://img88.imageshack.us/img88/7172/op132m03a2jw.jpg

http://img88.imageshack.us/img88/1506/op132m035wx.jpg

http://img409.imageshack.us/img409/8...46tg32m1qg.jpg

are they similar to Ballistix ?

Very nice work man, I wish I had 100 sets laying around lol.

Quote:

---

Originally Posted by Supertim0r

are they similar to Ballistix ?

---

basically same thing.:) but not 5B-D

Although, this is nice to see, on my book its a dollar late and a nickle short. Why? Because after 3 emails to Team rma, I never got a tracking number, because my IDE 0/1 is dead, and because I spent 270+ 15$ shipping (first time around) + 15$ shipping to send dead sticks back to tank guys, plus another 15$ shipping for some new sticks. So in effect, 350$ (plus 160$ for a newmobo) for 270$ pair of sticks.


If anything, im happy I got to purchase through TG, they have excellent customer service and responsability, and gave me an RMA # right away. As far as im concerned I wouldn't do Team ever again. :nono:

Quote:

---

Originally Posted by OmegaMerc

Although, this is nice to see, on my book its a dollar late and a nickle short. Why? Because after 3 emails to Team rma, I never got a tracking number, because my IDE 0/1 is dead, and because I spent 270+ 15$ shipping (first time around) + 15$ shipping to send dead sticks back to tank guys, plus another 15$ shipping for some new sticks. So in effect, 350$ (plus 160$ for a newmobo) for 270$ pair of sticks.


If anything, im happy I got to purchase through TG, they have excellent customer service and responsability, and gave me an RMA # right away. As far as im concerned I wouldn't do Team ever again. :nono:

---

Well I am sorry that it got you the unhappy experience , but I am also wondering how hard you try to figure out how to tune this ram and figure it out:)
I don't mean to argue with you or whatever your action against Team Group, but basically you seems to consuming more effort on researching all the "wrong idea threads here" against team group than what you need to figure out.
If you prefer Team group refund all your shipping cost, I am extremely happy to get all your Money back on that for your shipping lost.

Asus 939 board sux with memory compared to DFI
I had an A8N-SLi premium : 280ish 2.5-4-4-8
The same memory (OCZ 4800EE) with an Ultra-D : 300 2.5-3-3-6 / 315 2.5-4-3-6 with only 2.6-2.7v

I had an A8N32-SLi DX last week : 255ish with my ballistix (tried everything)
The same ballistix (bought new 2 weeks ago) with a new Ultra-D AD0 : 290 3-3-3-8 2.6v, 295 2.7v, 305 1m 2.75v

DFI rox...the A8N32 was really bad with my balllistix

Quote:

---

Originally Posted by Onepagebook

1.If you are Opteron user, please DO NOT use divider on running 5bf IC. Why, because OPty is well-known
lack of divider of good results.

---

eh...doesn't that kinda defeat the purpose of getting an Opty?
it kinda sucks to have your core clock limited by memory.
if you have a 148 it's less of a problem I guess because of the 11 multi but if I bought an opty I would probably be using a divider and I would not buy memory that's performance would go to shiit on a divider, esp on 250+ dollar memory.

Quote:

---

Originally Posted by snowwie

eh...doesn't that kinda defeat the purpose of getting an Opty?
it kinda sucks to have your core clock limited by memory.
if you have a 148 it's less of a problem I guess because of the 11 multi but if I bought an opty I would probably be using a divider and I would not buy memory that's performance would go to shiit on a divider, esp on 250+ dollar memory.

---

Exactly, but you gotta think about where the problem is. Regular a64 even 3000+ up to 4000+, you can say all the a64 line, don't have any single divider problem. And it's not only on this sepcific IC, it's on the CPU.:)
I can't even get 300x10 with 166 divider boot into windows on my 146.:slap:

Kev, you mean on your CABNE 0530AMPW? i've got one that's pretty similar, i normally don't use ratios, but just because you brought it up, i'm gonna look into it. i know that i've had to do silly things with MAL in the past when using a ratio other than 200mhz :stick:

well, damn, I almost forgot the 183 divider on that,
looks like no problem so far on running ddr540 same setting on 1:1
Good news..

http://img331.imageshack.us/img331/7...183d32m3il.jpg

http://img404.imageshack.us/img404/2...pi32m024jb.jpg

update:

opteron146 with 183 divider@ htt301 super pi mod 32m passed

http://img164.imageshack.us/img164/2...pi32m034fx.jpg

for my venice all I gotta do is move MAL from 7 to 8 and viola!, non-200 memclocks are fine.

glad I kept my venice and didn't go for one of them optys, the double internal cache isn't that big a deal. what is a big deal is a good mem controller, but I guess you need an FX for that.

opteron146 with 183 divider@ htt301 super pi mod 32m passed
windows screen captured:

http://img348.imageshack.us/img348/8...3pi32mp0os.jpg

Kev Thanks a ton for the tip on using a divider. I know my cpu has more in it but I have to run a 133 divider for my ram but it has a hard time clocking. I will mess around with Mal/Rp if I have time today.

Quote:

---

Originally Posted by OmegaMerc

Although, this is nice to see, on my book its a dollar late and a nickle short. Why? Because after 3 emails to Team rma, I never got a tracking number, because my IDE 0/1 is dead, and because I spent 270+ 15$ shipping (first time around) + 15$ shipping to send dead sticks back to tank guys, plus another 15$ shipping for some new sticks. So in effect, 350$ (plus 160$ for a newmobo) for 270$ pair of sticks.


If anything, im happy I got to purchase through TG, they have excellent customer service and responsability, and gave me an RMA # right away. As far as im concerned I wouldn't do Team ever again. :nono:

---

What does this have to do with this thread? Yet another complaint from you Omega. I am not convinced yet that you aren't wasting more than just your money, like possibly your time spent overclocking.

Thanks for your time invested in this Kevin. I am sure many of us are happy to see your input. Once I get my sticks I will be sure I am ready to use them versus just popping them in and crossing my fingers.

Thanks Ron:), I have talked to Omega , things look good so far, and pretty appreciate his understanding for giving his unbias try one more chance.

Quote:

---

Originally Posted by Onepagebook

Thanks Ron:), I have talked to Omega , things look good so far, and pretty appreciate his understanding for giving his unbias try one more chance.

---

No problem Kevin.

I have no problem defending something (or someone) that is being beaten down or belittled in a way that I consider unfair and biased.

I am glad you and Omega are working things out. Hopefully he will be more informed on the particular sticks he is getting (if that is what's going down), and how they perform with his particular motherboard/cpu.

well this could almost be used as a basic how to overclock 5B-F.. rather nice...

Quote:

---

Originally Posted by Onepagebook

Good points:

Answer are:
1.that date code you see merely stamped as Team Group package dated not Micron manufactured date. I can just tell in public;) all Team Group's IC date code is 0538 no more.

---

The dat code of my set is 0608.(On the ram ICs)

I got them from rma and asked them do not attach the heatspreader on my set.(In this way,I could see the date code on the ram ICs)

I think the date code is the key point of high FSB.Maybe those people who can't OC their ram well just get a set with other date codes on their ram ICs than 0538.

can we have a picture of that?

Are you testing all these with just single cores? Any Dual Core testing?

The reason I ask is because everyone knows dual cores are extremely power hungry, and it could be that the effects of this stress the motherboard even more. Who knows? Just my random thought :)

I really hope that I can get 1 pair of this here in HK, pretested by OPB.

Quote:

---

Originally Posted by ahmad

Are you testing all these with just single cores? Any Dual Core testing?

The reason I ask is because everyone knows dual cores are extremely power hungry, and it could be that the effects of this stress the motherboard even more. Who knows? Just my random thought :)

---

Sure why not? lol; Users' advice always is the goodness for listening.
I will do it on the fx60+ 7900gtx;) the most heavy power consumption;)

I test those on single core is because someone pm me :
OPB , would you please test on the single core, I think most of XS users are single core users;)"

Haha!

Well dont be testing all 100 kits with a dc, just one or two high clocking ones will do for us I think :)

Thanks OPB.

ahmad, here you go, I think this will at least do it quite good:)
completely original cooling for every parts on board;)

http://img101.imageshack.us/img101/2...5bf3d054fn.jpg

Perfect. Timings are very close to the single core results you have there, and this is stable in 3D. Good stuff OPB. Thanks :)

Quote:

---

Originally Posted by ahmad

Perfect. Timings are very close to the single core results you have there, and this is stable in 3D. Good stuff OPB. Thanks :)

---

Now , for those who got a bit confused by the memory parameters setting from all varities dfi nf4 boards, the 329 beta bios has been released and synchronize all same dram related code.

Quote:

---

Originally Posted by Onepagebook

Now , for those who got a bit confused by the memory parameters setting from all varities dfi nf4 boards, the 329 beta bios has been released and synchronize all same dram related code.

---

though i've yet to notice a difference in performance, though two things seem to be majorly different:
data drive level 4 is back
auto MAL/RP are much looser than before :confused:

ya that's why it said "synchronize":D
cuz ultra-D is way tunning for TCCD too much,
but I thing the DS/DDS concept is still there

What drive level for this ram on a regular sli-dr with the 329 bios? 8/4?

hey opb, first off, nice work... good way to redeem team's credibility... keep it up :cheers:

second off, i am curious what PSU you are using... is it something that most users have like a powerstream 520 or is it your big zippy? im thinking you can ensure us getting the results you get if you use a powersupply that we use ya know?

Quote:

---

Originally Posted by xgman

What drive level for this ram on a regular sli-dr with the 329 bios? 8/4?

---

5/4 is working amazing on my Ultra-d with my Crucials 300mhz 3-3-3-8 prime stabile.

http://img328.imageshack.us/my.php?i...lsprime8wt.jpg

Quote:

---

Originally Posted by Ferry82

5/4 is working amazing on my Ultra-d with my Crucials 300mhz 3-3-3-8 prime stabile.

http://img328.imageshack.us/my.php?i...lsprime8wt.jpg

---

keep in mind that the ballistix are 5b D, and this is 5b F. similar, but not exactly the same :toast:
nice clocks btw, i'm jealous :p:

mark, here is the answer you expect;)
this one also not bad huh?:D

I will have completely new thread to explain the settings and details;)
took from finished:)

http://img410.imageshack.us/img410/8...32mfd026sn.jpg

Windows capatured:

http://img323.imageshack.us/img323/2...mpi32mf3dy.jpg

Quote:

---

Originally Posted by Onepagebook

mark, here is the answer you expect;)
this one also not bad huh?:D

---

2mhz higher than me! :slapass: :stick:

:D:D:D

looks awesome! i'll keep my eye out for that thread, i'm curious to see what you have to say about 4x1gb :toast:

Quote:

---

Originally Posted by Revv23

hey opb, first off, nice work... good way to redeem team's credibility... keep it up :cheers:

second off, i am curious what PSU you are using... is it something that most users have like a powerstream 520 or is it your big zippy? im thinking you can ensure us getting the results you get if you use a powersupply that we use ya know?

---

Zippy 950 iirc, says something in the first post about it as well, iirc.

Quote:

---

Originally Posted by STEvil

Zippy 950 iirc, says something in the first post about it as well, iirc.

---

in case anyone cares, all the stuff i've done is on a powerstream 420 ;)

Any idea if/when these will hit Europe or the UK? I keep on looking and coming up with a blank. :(

I suppose I could get G.Skill HV, but I think these Cronus are better - and they would replace 5B-D (Z503 Ballistix). I just lost a whole night's Rosetta crunching to mem errors with those...

Quote:

---

Originally Posted by stone_cold_Jimi

Any idea if/when these will hit Europe or the UK? I keep on looking and coming up with a blank. :(

I suppose I could get G.Skill HV, but I think these Cronus are better - and they would replace 5B-D (Z503 Ballistix). I just lost a whole night's Rosetta crunching to mem errors with those...

---

PM me about details, thanks:toast:

@OPB
My Team Cronus (Micron) is also having some problems above 265, my mobo is an MSI Diamond Plus, can you help me here?

very good info thanks

On a Ultra-D would the old 704-2BTA bios be good for this memory or should I use the newest? I only just got my two sticks of this ram and havn't had much time to play, but stock speed (250 3-3-3-8) seems solid, but if i bump it up to 255 prime/spi fail fairly quickly.
Need to play around some more with timings as suggested in this thread, just wanting to know if it is futile using 704-2BTA :)

nice thread bump :p:

personally, i liked 11-14 the most for 5B F.. though 255mhz is very weak. let's see what your a64tweaker settings are. what voltage? active cooling? which slots? :D

oops didnt notice the date on the last post, just searched on stuff for my new ram :D

voltage i set to 2.7v which i think gives around 2.72ish in windows. No real active cooling, just got rear case fans closeish to where the ram is, and also a top exhaust fan, but don't have a fan purely to blow over the ram.

Using yellow slots, not sure which are meant to be better. Will get a64tweaker screenie up when i get home and will look into some newer bios'

ok here is my a64 tweaker screenie

regardless of if its 2.6v+.03, 2.7v or 2.7v+.03, tref of 3120/3684/4708, and couple other minor tweaks i cant get 255 stable. Prime will fail within 30 seconds, however at 250, any of the above settings will work... Guess i need to try a newer bios...
DS/DDS is 5/2 btw
Attachment 51041

all right, use the orange slots. ;)

tref 3072 iirc.. it's down near the bottom of the list :p:

everything else doesn't look.. too bad. a little loose, yes, but not too bad. if orange slots don't help a whole lot, check out the temperature of the sticks while running memtest. if they feel warm, seriously consider rigging a fan up over the memory. here's what i did:

http://eclipseoc.com/phpwcms_filestorage/1/1_117.jpg

and then perhaps try these timings?

http://eclipseoc.com/phpwcms_filestorage/1/1_217.png

you also *may* want a drive strength of level 7 or 8

TRFC 12 or 13 for best results. TREF 4706. Idle Cycle 16clocks. Dynamic Idle Counter...try enabled, but micron may need it disabled for any sort of high frequency, but try.

I wish the area around my RAM/CPU was that clear!

cheers for the replies guys, will try it out tonight :)

edit: so 11/14 and 3/29 bios' are best? might try them... that or 4/06

OK have updated to official 3/29 bios and moved ram to orange slots, and using 2.7v (2.72 real)

No luck with 260, even though i was able to run through a few loops of memtest86. Prime still fails within seconds at 260.

Have been trying settings (or slight variations of) from OPBs first post of this thread and also in this thread http://www.xtremesystems.org/forums/...ad.php?t=98014

Currently testing 255mhz with settings exactly as in the first post of this thread (for OPBs 265mhz screenie) so will see how it goes.

The RAM doesnt feel especially warm, but if i cant make any progress i guess il try to somehow rig up a fan.

Also despite OPBs settings saying use DRAM Data Drive strength of 2 or 3, when i set it to auto it uses 1 and if i set 2 i get either instant bluescreen upon loading windows, or windows is really screwed up (thought i'd totally corrupted the install)

edit: 255 seemed stable (3hrs prime) so upped it to 257.5mhz and the bastards failed prime after 2.5hrs at ~2.72v. Enabled the +.03v setting so ~2.75v now and has been priming for almost 3.5hrs. Now onto 260!