Printable View
It's been delivered to my apt, and I am still at work. I am a Art dealer in NY. I sell original paintings by Chagall, Monet,Picasso, Miro, Leger, etc. I have a client looking for a nice Monet, so I need to work, so I can make more money to buy ram from OCZ. at $360 a gig for ram with heatspreaders that fall off before you get them. I honestly don't care if they come with diamonds on the heatspreaders. I just want to see some performance and no bULL$H!t!!Quote:
Originally Posted by TEDY
You a bit confused Ryan I never said it was right or wrong and being a person who has never sent somthing back on RMA because of my damage (Honest). After reading my post i could see how it could have been taken the wrong way so i wont hold it aganist you accusing me of doing such. My statement's were based on sarcasm mainly because of alot of posts i see here and on other forums of ppl just saying I RMA'd it.
I have two MB's here if i was that type I could send back and RMA but i fubared em so i keep em. Hell i have only RMA'd 3 thing in my entire life! Memory to OCZ (you handled my RMA), memory to Muskin, and a MB to Asus.
Take some pepto and relax cause you worked up for nothing. It hard to express tone or sarcasm while typing.
LonG
Quote:
Originally Posted by ryanpgroovy
A little follow-up here. I just want to let you guys know, OCZ is being very proactive and responsive in addressing the bent spreaders. There is a lot to be said about a vendor that is willing to go to lengths such as this to ensure satisfaction.
Results so far. 200 MHz 2-2-2-5 worked fine. I got 2 test 5 errors at 210 MHz 2.75V same timings and just decided to go for the higher fsb's since my BH-5 should be what I rely on for 2-2-2-5 any way. So, with no CodeRed tweaking whatsoever, but with rated voltage and rated timings, these have passed rated speed no problem. (2.5-3-3-8, 2.75V, DDR533) Not earth shattering, I know, but I have seen enough TCCD users not hit ratings, so I thought I should post. I'll be going for top 2.5-3-3 speeds first, and later I'll start tightening/tweaking them up, so more to come. I'll keep the screenies smaller after this one though.
http://mysite.verizon.net/res6otqr/s...redBW_7925.jpg
Ah, and a question for Steve, Tony, or Ryan. I haven't used past 2.75V yet, but I want to clarify. 3.1V still keeps these in warranty, correct? I don't want to blow that until I know I'm happy with the memory.
My 4200EL plat spreaders look fine here, but I've seen spreaders come loose on Mushkin, Corsair, and OCZ. It happens. It's just thermal tape afterall.
On another note, I don't know of any other company that would give the kind of support that OCZ is showing right here on this forum as well as when you call them up.
Juice em up m8...see what they do ;)Quote:
Originally Posted by mikead_99
if onyl they would have been cheaper...360$ really out of my limit :(
Quote:
Originally Posted by TEDY
You must of missed it...
http://www.xtremesystems.org/forums/...057#post594057
~~~~~
On to that Q&A issue. I am strongly considering purchasing this RAM, have been following this thread and hitting all the resellers looking at price trends.
That heatspreader contact sitch doesn't impress me neither, too much, especially considering we are being asked to pay top $$. I'm not :banana::banana::banana::banana::banana:ing at you OCZ support guys, I think you are the best. I'm just wondering if you could convey to the people who are resposible for QC what effect that has on customers and potential customers. Maybe they could be made to see that it is probably wise to keep tighter controls on it. Like perhaps a visual inspection...
Word ^ :)
Scott
Quote:
Originally Posted by ryanpgroovy
I am sure theres a way for you guys to tell if he voided the warrenty. youd be able to tell if his ram was overvolted by the burn marks=no rma. You could implement measures that tell if you overclock or overvolt, maybe a tiny rom chip that writes a record if it senses more than 2.9 volts or more than 5MHz overclock. If someone rmas it, youd know.
Overvolt check is easy, and relatively cheap, and probably done by some manufacturers I'd bet. Fuse-able link, it's a trace that'll melt from exposure to a pre-determined voltage. Put it in parallel to the supply voltage trace and you can tell it's been overvolted without making the RAM stop working. Anything beyond that would be more costly and not as relevant any way. Going over-freq really shouldn't break anything in most cases.Quote:
Originally Posted by Geforce4ti4200
BTW, testing out your 2-3-3 @ 280MHz right now. Looks good so far in memtest86+. I jumped from 275 straight to 285 and couldn't find a happy place. I could reduce errors by increasing voltage, but at 3V I started getting more again. I saw you wanted 280MHz 2-3-3, and tentatively I'm doing that with 2.85V, but I'll know a little more when I boot Windows and test there.
275fsb is nice but id like to see it get over 250fsb dual channel on s939 ;)
You missed something. I upgraded, that is s939. :DQuote:
Originally Posted by Geforce4ti4200
edit to add: 280 MHz seems rock solid with 2.9V, with some moderate testing. Need to really give 'em some long term stress testing, but they're going to leave for new spreaders Monday and I think I'd rather just play over the weekend.
http://mysite.verizon.net/res6otqr/s...redBW_8330.jpg
Quote:
Originally Posted by Geforce4ti4200
both the 512x2 3700elplat and these 4200elplats do well over 250 dual ch 2.5-3-3-7-1T. First boot up with the 4200's at 265 and plenty more room.
Wow Mikead, I'm glad to see you are having good luck so far with the 4200. I have spent 3 hrs with memtest86 and I'm starting to lose it. I drank 3 redbulls and 4 cups of ice coffee and I am a little jittery after running over 200 different configurations of test 5. If I have to press (c-2-5-5-enter enter) 1 more time I'm going to kill myself.
Same as you, my heatspreaders weren't making contact with the ic's. I put my xray goggles on and found the ics are from 0440. this is pretty new stuff. I managed to fix the heatspreaders after a few minutes and they are fine now. someone at ocz put the modules in too high, so the heatspreaders weren't closing on it properly.
So far I've been unable to run these at thier rated speeds(2x512) no matter how many volts or 1t/2t command. I swapped cpu's from 3200cg mobile 62w to 3200+ CO claw(same results) However I tested each stick separately and found 1 likes volts around 2.7v and the other likes high volts 3.1-3.3. They both crap out at 280mhz 2.5-3-3-X 1T. I was able to run 1x512 @280 but even at 285 cas3 I was still getting errors. I've tried everything from Mr. Wu's tccd settings to trying my own settings, even auto. drive strength 2 or 4 make no difference. It funny how piccky this ram is with volts. You change one timing and the volts change drastically. it could make a differce of just .1v to make it or break it. Drive strenth 1 no boot. even 300htt 3-4-4-8 2t no go. I am sick. I don't know if a dual channel cpu would help at all. I don't have a 939 to test on, so I am going to do my best with a dfi LP-UT and 2 different cpu's. But so far this ram has been a major let down. $360 could've bought me a 939 and a decent 939 board. The highest htt at 1:1 was with some trusty EB3700 2x512 @3-3-2-10 1T @ 273mhz and that was with my mobile. I don't think this is my mem controllers fault, but the problem may lie generally within the 754 mem controller not being able to handle 2x512 well at high htt. Does anyone have a 754 and 939? Has your 939 been able to run 2x512 better than the 754? Steve or Ryan? Is there a way I can send these back to you for some sticks that run together at the same volts? Sean already got me an rma. And don't say I voided my waranty, because I had to fix something your guys didn't do correctly. (edit-nevermind, it's 3M's fault due to bad thermal tape) cough.. (If I sound like I'm being a jerk, I am just in a bad mood form all this memtesting. Don't take it personally.) :)
As far as the spreaders, I spoke too soon, my 4200's were completely coming off also. I spoke with Ryan and he said they just discovered a bad batch of 3M tape (he thinks) which was used on these. He will replace them or the spreaders if need be. I just popped mine off and with a fan over thgem and a booster they are fine. The samsung chips really are made to take a great deal of heat without any ill effect and really don't need the spreaders, and the booster fits better anyway.
(by the way, my 3700ELplats which I got over a month ago have the spreaders perfectly stuck on as do some earlier 3700EB's which will now go up fro sale.)
Quote:
Originally Posted by trans am
I don't have a 754, but I can telll you that it really is differnt board to board even in 939. With the Asus A8V, it couldn't get very high htt at all. Maybe maxing in the 250's whereas the Neo2 goes over 300.
Dang trans am, sorry to see you having such bad luck. I've seen you in a few threads now digging through the TCCDs and thought these might be the best bet for you. It is good to see you guys reporting your findings now too though. It's kind of hard to decide what to do with these based on just my results. 280 is good, and I'm happy with it, but I have seen a few dual channel TCCD setups get a bit more. I hate to gamble, but I would go for another set if I felt I had a decent shot.
Actually, it's C-2-5-5-ENTER-0. :DQuote:
Originally Posted by trans am
I totally know how you feel. This is why I picked up my 939 this morning. For $310, I'm now running dual-channel at 275MHz. I know these Platinums like voltage because they scaled well on my DFI 754, but the Neo2 limits VDIMM to 2.85V. I'm sure I can hit somewhere in the neighborhood of 290MHz to 300MHz if I had more voltage.
Go 939. Watch your RAM shine. ;)
Last few days... i pressed that combo of key so frequently that I didn't recognize it by seeing it here :P Was wondering for a second there.. hehQuote:
Originally Posted by conrad.maranan
I am still looking forawrd to see dual channel result on 939 with this memory :P
Booster 3.35v :stick:Quote:
Originally Posted by conrad.maranan
I don't have a Booster yet, as I just purchased this board and CPU this morning. :stick: I was hoping that VDIMM options would at least extend to 3.1V.Quote:
Originally Posted by xgman
Man guys, If you noticed I was offline, it ws because I was testing more. I came to the conclusion that this pair won't work together no matter what setting. Even at ddr400 2-2-2-6 2t 1.5-3.4v and everything in between there is always errors in test5. and I don't mean 2 or 5 errors, we're talking at least 5,000+ errors!!!! I tried resetting cmos, changing pots on powerstream(thought maybe rails were too high or too low) tried slots 2 and 3, 1 and 2 and 1 and 3. no matter what, nothing works. 2 different cpu's=same Fvcking$H!T!!!! Ocz has never steered me wrong in the past, but even after spending $917+ dollars on ram this month my quest for good tccd goes on.
So far:
Mushkin ver.2 blk tccd=crap
Kingston KHX tccd=crap
Ocz el4200platinum=ironically the most expensive and was the worst of the 3 sets tested. Hopefully when you guys @OCZ get my ram(btw heatspreaders are coming off the ic's again.) you can give me your best tccd yet so I can come back to XS and show my fellow members that ocz really is the best. I know that accidents happen I am a very understanding nice guy but you can see why this looks really bad. :(
Wow! :slobber: That's a $#!tload of money on RAM...and in one month, too!Quote:
Originally Posted by trans am
I'll be home in about an hour. I'll upload my screenshots when I get home. :)Quote:
Originally Posted by jinu117
C-2-5-5-enter-enter also works. Trust me on this one. :(Quote:
Originally Posted by conrad.maranan
My guess is that the set isn't volt matched. Speed matched, yes, but each stick likes completely different volts. 1 loves low volts around 2.7-2.8 the other one only works with volts of 2.9 or higher. If you give the stick that doesn't like volts too much it gets errors. If you don't give the stick that likes volts enough, it fails too. So this is why they won't work together. Now if I could only have 2 sticks that run at 280 or higher with the same volts I'll be happy. :)
Quote:
Originally Posted by mikead_99
we did not know this! what cpu and how does it overclock? has it scored higher than your 3700+ like it should? Can you reduce the voltage to 2.7 and see how they do then? Cause if I am right, the gigabyte for s939 had 2.7x max? :confused: 280x9=2.52GHz with my winchester in 1:1 would be real sweet. I really hope my corsair can even do 5:4, if not ill be forced to upgrade since running 3:2 or lowering the cpu clocks will just cancel out and itll be a sidegrade to my newcastle.
"but the problem may lie generally within the 754 mem controller not being able to handle 2x512 well at high htt."
your lucky, mine cant even handle over 250fsb even with a single 256mb or 512mb stick, this is why I could only get 250fsb out of my corsair value!
depending how well my corsair value does, I may or may not need tccd ;)