safe micron DDR3 voltage

[TheHardCase]

Not sure if there is a problem ect.

I've been running straight for a good 16 hours, thing is I went to lower timmings & volts , no go.
It will not run anymore at default speed, still works at overclock posted above.
Any Idea?

[dinos22]

Interesting...
But I must admit, isnt 72 quite high in first place? I've been running 45 trfc DDR3-2000???

In my experience: ras to CAS is the true sinner when it comes to increasing voltage...

it is
for benching the lower the better as long as you don't sacrifice frequency too much heh

i seem to be at 56 for 2GHz+ on nvidia with 8-7-6-20 1T 1.9-1.96v

[M.Beier]

it is
for benching the lower the better as long as you don't sacrifice frequency too much heh

i seem to be at 56 for 2GHz+ on nvidia with 8-7-6-20 1T 1.9-1.96v

Well in my case... I just found maximum frequency bench stable, then I lowered from 60 to 42 without any trouble...
Think it was 985mhz CL6-5-5-18 PL6 or so...

[KTE]

Thanks saaya, I think it'll help many many users a lot as DDR3 is completely new territory and this forum is not just read by xtreme hardware killers but also mainly by xtreme+safe users in general

I remember coming back online after roughly 18 months and being new to DDR2 completely while it was quite fully accepted by then (courtsey of D9) and I know what it feels like to hunt for some honest and reliable advice within so many opinions and high volt usages. Luckily, reading where to start off with 24/7 oc settings and not risk deaths, I took Enjoys advice for volts and trusted that and he was bang on. More DDR2 experience and my findings were the same as his for D9 voltage

[dinos22]

^^^ True that ^^^

[saaya]

Not sure if there is a problem ect.

I've been running straight for a good 16 hours, thing is I went to lower timmings & volts , no go.
It will not run anymore at default speed, still works at overclock posted above.
Any Idea?

hmmmm on the 790 or also on another board?
did you chek with ocz if there is a new spd maybe or some recommended memory settings etc?

So at stock it doesnt work, but oced it does?
Well maybe it degraded and now requires more volts to work stable?
That sounds more like ddr degrading though, i never heard of this from ddr3 so far. Are you 100% sure it was stable at stock speed on this board to begin with?

dinos and m.beier, the guy who posted here didnt ask for tweaking and benching advice, he wants his mem stable, not tweaked, thats why ryder recommended higher trfc, which is you can recommend really not knowing the details of his setup and what the problem could be

Thanks saaya, I think it'll help many many users a lot as DDR3 is completely new territory and this forum is not just read by xtreme hardware killers but also mainly by xtreme+safe users in general

if this thread prevents a single person from blowing his mem im happy and the effort was worth it

I remember coming back online after roughly 18 months and being new to DDR2 completely while it was quite fully accepted by then (courtsey of D9) and I know what it feels like to hunt for some honest and reliable advice within so many opinions and high volt usages. Luckily, reading where to start off with 24/7 oc settings and not risk deaths, I took Enjoys advice for volts and trusted that and he was bang on. More DDR2 experience and my findings were the same as his for D9 voltage

yeah, there are many people who post results to boost their ego and the shared information is not really useful for most people who dont want the highest scores no matter what

[Achill3uS]

hmm, guys what about this...
OCZ DDR3 1600 7-7-7 Platinum kits, clocks timings doesn't matter could be at stock, but as i feed it with 1.95v+ become more unstable, and got more errors in memtest 2.01
it seems my P5K3 Deluxe absolutely hates this memorys, but in Striker 2 Extreme clocks nice, 1600 6-6-6 1.94v, 1780 7-7-7 1.94v pass
maybe d9jnl?

[saaya]

well, meassure real vdimm
all micron ddr3 chips have a certain voltage they like, above that they give you errors and are unstable. with high vdimm you can get nice screenshots at high speeds and maybe some light benchmarks, but itll never be 100% stable.

[M.Beier]

well, meassure real vdimm
all micron ddr3 chips have a certain voltage they like, above that they give you errors and are unstable. with high vdimm you can get nice screenshots at high speeds and maybe some light benchmarks, but itll never be 100% stable.

Did you ask Sam?
2.4v is quite high voltage for DDR3, atleast in my humble oppinion... Torture testing... P95 or memtest, I dont know - but I'll call it stable either way... And 1300 hours is = 54 days, so if I've got the right information.... I'd say your pretty much wrong...

[Achill3uS]

well, meassure real vdimm
all micron ddr3 chips have a certain voltage they like, above that they give you errors and are unstable. with high vdimm you can get nice screenshots at high speeds and maybe some light benchmarks, but itll never be 100% stable.

i will measure it if i got home
that's okay
BUT
with this memory till i can't reach 1800+ 7-7-7 / 7-6-5 stable
wich speed was memtest stable on p5k3 deluxe (with cellshock 1800) around 2.0v, and could easy boot and bench 1000mhz 8-7-6 and on S2E still can't boot over 1900mhz mem speed, so i need more experience with 790i or this OCZ's are weak (and has jnl wich - it seems - don't like higher voltage 24/7) - maybe both

[TheHardCase]

[QUOTE=saaya;2892472]hmmmm on the 790 or also on another board?
did you chek with ocz if there is a new spd maybe or some recommended memory settings etc?

So at stock it doesnt work, but oced it does?
Well maybe it degraded and now requires more volts to work stable?
That sounds more like ddr degrading though, i never heard of this from ddr3 so far. Are you 100% sure it was stable at stock speed on this board to begin with?

Correct, at the time this happened I also lost vista 64 as posted to the other thread. It ran before for hours at stock & will not no matter what I try.
100% stable on overclock for now anyways at 7-6-5-18 @ 1T 2.06 VDimm real.
1800MHz 1:2 noproblem, but not stock 1600MHz 7-7-7-18 @ 2T

P.S. this is not ocz, it's viper patriot PVS32G1600LLK

[BadNizze]

2,16v for 1 houre here on P5E3 Premium, no damage or degradtion. Anyone know messure points for my board? Based of this tread i will not go over 2,2 for benching with my STT ProdjectX

[Kasparz]

hmm, guys what about this...
OCZ DDR3 1600 7-7-7 Platinum kits, clocks timings doesn't matter could be at stock, but as i feed it with 1.95v+ become more unstable, and got more errors in memtest 2.01
it seems my P5K3 Deluxe absolutely hates this memorys, but in Striker 2 Extreme clocks nice, 1600 6-6-6 1.94v, 1780 7-7-7 1.94v pass
maybe d9jnl?

D9JNL. Yes, over 2.0Vdimm gives instability and errors.

[saaya]

Correct, at the time this happened I also lost vista 64 as posted to the other thread. It ran before for hours at stock & will not no matter what I try.
100% stable on overclock for now anyways at 7-6-5-18 @ 1T 2.06 VDimm real.
1800MHz 1:2 noproblem, but not stock 1600MHz 7-7-7-18 @ 2T

P.S. this is not ocz, it's viper patriot PVS32G1600LLK

alright... so it doesnt work stable at 1600, even with increased vdimm?
then its def some fsb/mem divider issue!
this is exactly what happened on 680 too, so it would make sense...

a certain fsb/mem combo would work stable and then later it wouldnt for some reason.

2,16v for 1 houre here on P5E3 Premium, no damage or degradtion. Anyone know messure points for my board? Based of this tread i will not go over 2,2 for benching with my STT ProdjectX

there should be marks on the pcb between or next to the dimm slots.
round circles where extra caps should have been placed. in those circles there are 2 dots, one is round one is quare. meassure the square dot vs ground and its pretty much vdimm. not sure its the best spot to meassure but they easiest to describe especially if not dont know the board layout of a board that well

[Stelios]

nice!
any pics?

I'll take some and post them !

If you can remember ... you saw this modded dominator after the o/c party here in Athens (October 2007) . I had given it to hipro5 during the show . And to remind you who i am ... we got in the same car after the party finished and headed to hipro's lab . In fact ... we first went to eat something . BUT you began taking photographs of the food instead of eating it !

so... were you using only one stick to test more vdimm? or did you use both sticks, but only one degraded and the other was still ok even after 2.35v bios?

I started benching with both dimms installed . So , same voltage applied to both dimms .After a few minutes i noticed a strange instability , so i removed the dimms and tried them separately . One was still perfectly fine , while the other had degraded .

[saaya]

hahahah i see!
so your that big guy with the gangster sunglasses and the gun?
yeah i remember eating with you, your buddy and fredyama
thanks a lot again!

hmmmm so one dimm degraded, the other was fine...
do you remember which slot the module was in that degraded?
it was the one further away from the cpu right?

[Stelios]

hahahah i see!
so your that big guy with the gangster sunglasses and the gun?
yeah i remember eating with you, your buddy and fredyama
thanks a lot again!

hmmmm so one dimm degraded, the other was fine...
do you remember which slot the module was in that degraded?
it was the one further away from the cpu right?

I was the other guy ... not the driver .

The degraded module was installed in first white slot (closer to cpu) .

[saaya]

Ah ok! heheheh
first white slot hmmmm so closer to the cpu hmmmm

[Stelios]

Ah ok! heheheh
first white slot hmmmm so closer to the cpu hmmmm

What's wrong with the first white slot ? Why do you find it strange ???

Here are the pics of the modded dominator airflow with 3 delta fans you asked .

As you can see , the delta fan are taller from the original corsair used (28mm vs 20mm)

PS . Dust everywhere ... i have to clean it soon ...

Fan is FFB0412VHN : http://www.bravoelectro.com/pdf/FFB40X28.pdf
9500 rpm & 15.79cfm each one .

[Vladmax]

Hi, I am considering some DDR3 for my new 790i board and came across the sticks talked about here http://www.xtremesystems.org/forums/...55#post2897555

Does anyone know

1. what chips are on this particular RAM (Micron, Elpida etc.)? If it is Micron are they GTR or GTS?
2. And what is the difference between these two kits
http://www.patriotmemory.com/product...id= 634&type=1 and
http://www.patriotmemory.com/product...id= 670&type=1 except the heatspreaders and that one is a bit cheaper then the other?

Cheers

[saaya]

stelios, its interesting cause in my experience the second and 4th slot overclocked better than the first and the third.

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

[kiwi]

Links dont work right but any kit with 1600+ rating is more than likely to be gtr (jnl)

[saaya]

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.

[kiwi]

ok, 1600+ at least cas 7

Maybe there are some 1600 cas 8 or cas 9 kits that are not microns

[saaya]

actually samsung ddr3 can reach 1600 cas6
and micron isnt micron, dont forget he wants to know where he can find d9gtr/d9gts, not D9jnl or other die shrink micron ddr3 chips