Yap, great set I got. Hope it lasts (week18 of 09) :up:Quote:
Originally Posted by Yellowbeard
Printable View
Yap, great set I got. Hope it lasts (week18 of 09) :up:Quote:
Originally Posted by Yellowbeard
Thanks, I don't think anyone has been as pro-active as you guys in all of this Mike.Quote:
Originally Posted by Yellowbeard
Thanks a lot for clearing my doubts Yellow! Ok, it seems that I will put some basic memory and then buy a really great kit, wish this nightmare finish quickly so I can go directly to the Domis GT.
Thanks a lot Yellow and congratulations for the great work!
Currently I'm running my second TR3X6G1866C7GTF kit after the first failure.
oc @ 2000MHz 7-8-7-20-1T Vdimm @ 1.72v, qpi/vtt @ 1.65v, Vcpu @ 1.39v.
With the above settings I ran successfully hyper pi 32M, memtest for windows 500%, memtest86 5 passes stable and linx 5 passes stable.
Also I noticed that sometimes when I used lower vdimm values @ the post process only 4 giga where recognized from the 6 total. Rebooting raising vdimm and 6 giga recognized again. I remember another thread about dropping ram channels. I don't know if this is a bad sign of feature failure.
Using a dmm and measuring vdimm @ full load I noticed that the mobo undervolts the dimms by 20 mv.
chew, i experienced this over and over... thats the ceiling of the memory or the memory controller... whenever i pushed memory to the max when working for cellshock, regardless of what chip, it would always act like that when i hit the ceiling... itll not clock higher and vdimm wont matter and more actually makes it worse... if you make a graph you will see that it kinda bounces against the ceiling at some point, with increasing vdimm it goes into a sinus curve bouncing against the ceiling, or, it drops down bit by bit, or, it starts falling big time to very low speeds... ive seen all of those possibilities with the same memory chips from the same batch even...
but they all, always had one point at which more vdimm didnt help and started to make things worse.
btw, i had a strong suspicion back then that at least for micron chips pushing above vdimm that had them stable through the chips damaged the chips after some time. i always used memtest86 and kept vdimm at the level where it was stable for that speed and that timings, and those sticks never died for me... the only ones that did where the ones where i forced speeds higher and higher in windows with more and more vdimm, ignoring that the mem was actually unstable and just pushing more and more for suicide shots.
oh and yes, diferent timings mean more vdimm is possible or you get errors at the same vdimm that was ok before... in general in my experience tighter timings at the same speed = higher vdimm accepted by the chips. for micron cas and trp made a notable diference in how much vdimm i could pump through them, which then meant i could get slightly higher clocks... but those are tiny things, 4mhz more and 1 step higher vdimm...
corsair compatibility tends to be very good, but i dont think they referred to compatibility here, a couple of people seem to be unhappy with memory tweaking and overclocking on the evga board, not sure what exatly though as nobody mentioned any details...Quote:
Originally Posted by Yellowbeard
Add my Mushkin 6gb redline 2000mhz cl8 kit to the graveyard haha. Just died on me. The stick in DIMM1 just failed. I was actually looking at XS and it bsod'd on me and it has never happened with this kit, and it said something about anti-virus in the bsod, but I knew immediately what to look for since I've been through 15 kits of Kingston. This kit of Mushkin lasted much longer though. I've had it running strong for about 2 months with no problems, much different story than the Kingstons. Oh well, I guess, elpida just sinks farther into the hole they dug for themselves. First japan has to bail out elpida, now as if things weren't bad enough, the hyper chips are flawed and manufacturers are returning them in masses.
could you repeat that? it sounded like you had 15 kits kingston elpida hyper die on you? :eek: maybe its a typo? :oQuote:
Originally Posted by posershadow
about japan bailing out elpida... hah!
the irony goes deeper, japan bailed out elpida who then almost bailed out some taiwanese dram makers, but the taiwanese government put a stop to it from what i heard... funny eh?
help we need money to survive!
then take the money over seas and invest it in another company LOL :rolleyes:
just another reason why im against all those government bail outs...
i dont think this hurts elpida at all... the chip recalls are probably peanuts for them... what they really lost, and lost big time, is credibility...
i wouldnt have expected them to have so bad Q&A and then not even having the b4lls to stand up and admitting a problem... :rolleyes:
Quote:
Originally Posted by saaya
Absolutely true!!!!!!!:up:
oc @ 2000MHz 7-8-7-20-1T Vdimm @ 1.72v, qpi/vtt @ 1.65v, Vcpu @ 1.39v
how far of an OC does it take ???
at stock speeds and voltage are the sticks failing ???
Pushing ram speeds at low volts with ICs becoming very small I bet is not an easy task. Then people like us OC/abuse the sticks and then they snap and fail thus being the companies fault???? My CPU ever fails it will be due to my voltage and OC I put on the chip.
The company and the consumer should bear the responsiblity
Yes, some users have sticks dying using voltages that are below stock. Something is certainly amiss, perhaps the real rated voltage and speed of the IC's is a lot lower than what Elpida or the vendors applied, or simply issues in fabrication and quality control. All parts of the jigsaw I guess.Quote:
Originally Posted by iboomalot
Does anyone have a kit of Corsair 1866GT's (or 2000 C7 or C8's) that are week 26 of 2009? (IE: the code starts with 0926). If so how are they going?
This is the second kit TR3X6G1866C7GTF I'm using. With the first one I lost one module and everything was at stock. It's been a month now since I started using my second kit without any issue.Quote:
Originally Posted by iboomalot
No, at stock speeds and voltages the stick don't fail.
The above setting were the minimum I could use to keep the system stable and they are safe:http://www.xtremesystems.org/forums/...34&postcount=1
Corsair also in its forums suggests a qpi/vtt no more than 1.7-1.8 volts.
Paying such a high price you do expect more.
Of course everyone is responsible on what he does.
10 died on me, 5 were doa kits from newegg, probably the very same kits I was sending back. Only around 10 kits died in my hands, but if you looked at neweggs records, they probably would say that I killed all 15 haha. They never really understood what was going on. I kept telling them it was a common problem of the chips dying, but this was early on in the game, well before anyone on here had reported the hypers dying, so they thought I was just an idiot who didn't know how to install ram or adjust voltages. That's what you get for ordering 7 kits of the stuff haha. I couldn't have picked a worse set of ram to buy a lot of. One by one the dominoes fell and each kit had to be rma'd. Thankfully for my wallet, newegg paid the return shipping. Each kit was rma'd once, plus when I sent a kit in for refund, they sent me another doa kit back for no apparent reason which makes up the 15th set. Many hours of customer service later and I've got all my money back. Now it's on to the mushkin customer service department.Quote:
Originally Posted by saaya
damn you got me :rofl:Quote:
Originally Posted by G H Z
lolz i hope this kit doesnt die
they are good for 1120MHz 7-8-7-21-1T with nice subz at 1.76v :D
Your 100% correct , i also experienced the same situation with newegg, and was 1 of the first to buy these kits too.Quote:
Originally Posted by posershadow
Their policy is to resell the returns no matter what,you must of got mine :rofl:
I also wrote bad stick and do not resale on sticky note ,and taped it on mem stick,and i also called the supervisor to take them off the shelf.
Yeah, as soon as I started getting open packaged ram with tape sealing it shut, I started thinking hmmm... these look mighty familiar, then to test them and find 1 stick doa, I realized, aha newegg is just sending the junk ram back out. I even told the customer service rep online, I've gotten multiple packages that are open and doa, you guys are just sending the same ram from these rma's back out. Of course she ignored what I had to say and asked me what the symptoms of my computer problems were. Do you have the ram inserted right? Does your motherboard accept ddr3? Are they in the right slots? And every time the answer was the same, guys, I'm an overclocker, if anyone knows how to spot dead ram, it would be me.Quote:
Originally Posted by nge769
These chips were taken off the market for a reason ,they degrade.Quote:
Originally Posted by iboomalot
I,m using corsair 1866 cl7 3x2g..@ 1600, 7 7 7 20 1T 1.56 v because i don't wanna loose em. Hell, I'm afraid to look at them the wrong way..lol.
I wonder what they replace these chips with.
what about a KHX16000D3T1K3/6GX today? did kingston find out a solution for sudden deadh issue of his memory?
Same question here.
I own a Kingston HyperX T1 6GB and one of the sticks just give me Radnom BSODs and crashes, regardless of the speed (even at 1066 CAS8.)
What chips are we going to get in them if I will RMA the sticks ?
i'm rmaing my dom gt and i asked them which one they'll send back, i'm waiting for their answer
will let you guys know the answer
Add me to the list. One stick of my mushkin 998707 died last week. I was running it at 1800 7-7-7 1.65 so I wasn't pushing them hard. Still awaiting my rma #.
Posershadow: did you deal with newegg or mushkin for your rma ? If you dealt with mushkin ,what replacement set did you get ?
when did you bought them? before 1st of august?Quote:
Originally Posted by The-Fox
Well those original sets were kingston. That was all through newegg. My mushkin that just died is still in process with mushkin, I was late sending them in, getting lazy after having to go to the post office so much for all this elpida haha. No replacements yet. Hope they don't send me more bad elpida.Quote:
Originally Posted by azureoval
Yes, It was way before 1st of August.Quote:
Originally Posted by prznar1
Well one of my sticks went dead last Sunday. It was Kingston's.
Before it died I tried 1.48VTT and 1.72VDIMM along with Corsair C7-1600 Ver 2.2 for a 12GB 9-9-9-24 2000Mhz 1T Testing. After that it just died out.
What I've seen is Kingston has been replacing these kits for a new model. Let's hope that new verison will solve the issue permanently..
EDIT: Oh one more thing. That dead stick just won't boot up or recognize. However it will boot up just fine with my Rampage Extreme and I wonder what's wrong...