safe micron DDR3 voltage

[kiwi]

Any more info on those samsungs? Have you tested them?

Well, some D9JNL are certainly not bad

[Kasparz]

None of D9JNL is bad actually

[Vladmax]

vladmax, ask the manufacturer
but as far as i know nobody besides cellshock guarantees gtr/gts chips...

Manufacturer can not disclose such information... that was their answer...

[saaya]

Manufacturer can not disclose such information... that was their answer...

of course they can, they dont want to
several memory makers are playing a big poker game and dont want to play with open cards... which DOES make sense as there are some capy cats out there who will just do what they do and copy them, so it makes sense to keep it secret what chip your using... to some extent...

if a product has been out for a month or two i personally dont see any point in hiding what chips your using tho...

[Borgod]

not really, some ocz 1600 uses quimonda memory afaik, and there is some samsung ddr3 that can reach those speeds too...
and then there is the micron die shrink, which clocks good enough to make 1800 or even 2000 ddr3 out of it, yet its not as good as D9GTR when it comes to cas6 ocing.

You sure about that?

http://www.xtremesystems.org/forums/...&postcount=386

[M.Beier]

I just like to call:
BS

My arguement still havent been discussed... and fact is people are killing their ram with BELOW stock voltage...
- I'd say its not only vdimm... But hey, if you dont buy my arguement, keep your fancy ideas while I'll laugh my butt off...

PLEASE read my previous post and LOOK at whats being stated before replying to this post... Because obviously, evidence doesnt mean a jack@ss

[Charles Wirth]

1.5v killed ddr3?

link?

maybe a bad psu voltage spike or installing into a machine with the power on.

[M.Beier]

1.5v killed ddr3?

link?

maybe a bad psu voltage spike or installing into a machine with the power on.

http://www.xtremesystems.org/forums/...1&postcount=30

Im not talking JEDEC standarts, but what companies warrant... ;-)
I've only seen none micron at JEDEC volts...

[dinos22]

http://www.xtremesystems.org/forums/...1&postcount=30

Im not talking JEDEC standarts, but what companies warrant... ;-)
I've only seen none micron at JEDEC volts...

Beier
it's a one off man.
How can you be so certain that it's not just an odd one out or that it was dead from previous tests or whatever

1x example means little

[M.Beier]

Beier
it's a one off man.
How can you be so certain that it's not just an odd one out or that it was dead from previous tests or whatever

1x example means little

An odd one out... Thats a possibility, yes... dead from previous tests, because I know exactly how they're being tested before shipped to the stores, with the kit he has... And I have great confidence in that process...

I'd find it WAY more likely as I've stated that its not voltage, but a TIMING, a setting that injuires the ram...! Because if it was ram, and we're talking same chips here.... Dont you find it a little funny that someone injures their ram at 25-30% less voltage then others run without hassle? Well I know that I do... Further, from the DDR2 we saw same tendency, with timings, other factors having severe impact....

I think its completely BS to call "ohh its voltage"...

And from how I read the post I've linked to, I see 1.7v killing ram, GTR, I know people that has ran 2.8v without injuring their GTR... thats more then 60% higher voltage.... Now I just DONT belive that can be coincidence...

Ofcause the rate of electromigration is effected, yes, but thats not meassureable for us normal consumers... And I dont think the impact is that severe...

[dinos22]

Ofcause the rate of electromigration is effected, yes, but thats not meassureable for us normal consumers... And I dont think the impact is that severe...

i think this deserves some more scrutiny

there is no way anyone here can say with reasonably good confidence that any one rule applies in killing RAM >> unless you use extremes......hence likelihood or RAM dieing while running 2.6v constantly is much higher than sub 2v basically.........but these can only be expressed as general guidelines

you mentioned drive strength in other threads but we don't you tell us something we can understand. What are you talking about exactly here

[M.Beier]

Ihh, I'll explain later
I'll add some BIOS photoes, sadly Im sitting with my laptop... I could write the entire reason, actually I just did, but deleted it... Its just not usefull... (2 dead moboes, plus I've borrowed alot of HW to a friend thats hunting danish records with 2x HD3870X2)

I'll post pictures on saturday if I get the time... I wont be at my adress with the HW before then...
We (Oliver, THC, and myself) have 3x X3110, 1x QX9650 and want sub 8s 1M, PCMark2005 WR, plus 10 3Dmark WR's (x19xx + another class), so maybe I wont get the time, but I will try!

[dinos22]

Ihh, I'll explain later
I'll add some BIOS photoes, sadly Im sitting with my laptop... I could write the entire reason, actually I just did, but deleted it... Its just not usefull... (2 dead moboes, plus I've borrowed alot of HW to a friend thats hunting danish records with 2x HD3870X2)

I'll post pictures on saturday if I get the time... I wont be at my adress with the HW before then...
We (Oliver, THC, and myself) have 3x X3110, 1x QX9650 and want sub 8s 1M, PCMark2005 WR, plus 10 3Dmark WR's (x19xx + another class), so maybe I wont get the time, but I will try!

tell us what bios settings you are talking about
and we will do some thinking while you're breaking all those records

[BenchZowner]

He's talking about DRAM Drive Strength ( usually found in DFI BIOSes... )
It's a repeat of the conversation from the UTT-BH5/UTT-CH5 era again.

As a theory it was "ok", but in practice... people have killed chips with light & strong Drive Strengths and the same voltages... so-so.

[dinos22]

i agree with you BZ

Drive Strength in AMD days wasn't really as big a killer as some guys were making it out to be

[M.Beier]

He's talking about DRAM Drive Strength ( usually found in DFI BIOSes... )
It's a repeat of the conversation from the UTT-BH5/UTT-CH5 era again.

As a theory it was "ok", but in practice... people have killed chips with light & strong Drive Strengths and the same voltages... so-so.

That is correct... : - )
However, I dont think its one setting alone... I'll try to show once I'm diggin' down... However, back in the days DQS Skew + Drive str was a dangerous mix, however, Im proud to announce I've never killed a UTT/BH5 etc... Eventhough 4,03v ;-)

I have to get some sleep now, its 3:25AM CET, and I have lessons 9:10... Hope you guys will have me excused... : -)
- I'll check in late tomorrow, I have classes, and then squash + fitness...

[saaya]

You sure about that?

http://www.xtremesystems.org/forums/...&postcount=386

according to this thread:
http://www.forumdeluxx.de/forum/showthread.php?t=443947
ocz gold 1600 is with quimonda mem
i heard some other stuff too, but i dont want to speculate...
ask tony or ryderocz if you want to be sure...

I just like to call:
BS

My arguement still havent been discussed... and fact is people are killing their ram with BELOW stock voltage...
- I'd say its not only vdimm... But hey, if you dont buy my arguement, keep your fancy ideas while I'll laugh my butt off...

PLEASE read my previous post and LOOK at whats being stated before replying to this post... Because obviously, evidence doesnt mean a jack@ss

who are you referring to and what are you talking about?
and calm down man... geez... can you please stop posting bold statements and rants in this thread? its really annoying...

http://www.xtremesystems.org/forums/...1&postcount=30

Im not talking JEDEC standarts, but what companies warrant... ;-)
I've only seen none micron at JEDEC volts...

so what? now first you raise hell here claiming even 2.5v is perfectly fine for 24/7, now all the sudden your making bold claims and statements again this time claiming 1.7v can kill ddr3... you just proved my point, you love to be the center of attention and make bold claims and call everybody or everything wrong and bs... please STOP it, its annoying the hell out of everybody...

its not what you post but how you post it, that link was great, thanks for heads up, and yes, i agree, timings and drive strength DO have an impact on damaging memory, but geez, whats up with your attitude man?
just calm down...

cant you just for once make a calm and objective post about something without ranting, calling everybody and everything wrong and bs and claiming to have found some revolutionary formula or whatever?
if you want to go on doing this then PLEASE stop posting in my threads and make your own, THANK YOU!

now back on topic, as many of you noticed, micron ddr3 can take higher voltages, or, likes higher voltages and scales more, with tight timings.
with tightr timings the mem is working harder, and can suddenly take higher volts. at the same time with more relaxed timings the mem is unstable at lower volts, which makes you wonder if its just unstable, or if its actually getting hurt/damaged from the volts sooner if its running more relaxed timings. i dont know how to explain this... but the timings seem to have an impact on what voltages a kit can take. from what ive seen its a very small impact though and it wouldnt explain why 1.7v killed mem even if it was only at 1333 and 999... i suspect it was/is a bad stick, bad mobo, bad psu, or maybe ESD damaged the stick and then it failed immediatly or after some time in the board.

[BenchZowner]

You have a point there Saaya.
Oh, by the way: "Tight timings, high voltage & higher current draw"

I have yet to get past 2.29V on DDR3 ( and that's because I want to keep those 2 good kits that I have here alive ).

However...when I get... another kit....and the time to do some tests... I will be exploring the voltage limits

[Tony]

To ALL.

OCZ Gold is low to mid range ram, it may or may not be Micron based from us.

Platinum, Reaper and Flex are high end parts and feature high end IC's. All Platinum 1600 and higher is Micron based. Some Gold may have micron as some have seen as we shift so many IC's some just gets used up on lower end parts...there is no guarantee it will clock high though and you may get qimonda or elpida.

Also I hear talk of D9JNL, OCZ have been using this IC for 3 months now, its not new as some think, its been out a good while.

hope this clears a few doubters minds up

[Tony]

i agree with you BZ

Drive Strength in AMD days wasn't really as big a killer as some guys were making it out to be

I can't see how you can pass judgment on a subject like this as you have no idea about damaged quantity, only someone working for a module manufacturer would have any idea.

What you read on forums is around 1% of the total issues manufacturers see, the forum world is a small world on the whole. An example of how bad things get on forums is when i see posts about a new ram IC being good...you have to test hundreds of IC's across various week codes and die revisions to know whats good and what is not, just quoting what 1 end user managed to get from his or her modules does NOT tell the true story...trust me


There is sooooo much you guys are unaware of its untrue, issue is I can not tell you

[Tony]

have some info for you:

[dinos22]

good info in the graph

[kiwi]

Is there such DDR2 graph from micron?

[Tony]

Is there such DDR2 graph from micron?

this is for the newish DDr2 1066 cas7 part.



OCZ recommend a max of 2.2V and lower, most find 2.0V is all that is needed.

[eva2000]

new DDR2 part is that the stuff being used in most 2GB module DDR2 right now ?