Printable View
just got my twinmos bh-5's week 9 year 05.
smids did you go directly that high without any burn in? mine are kinda acting wierd. i uninstalled the nvidia ide software as i heard it's been giving people problems with overclocking so we'll see. i'll be posting by friday what i've gotten so far.
also can i do the burn in with both at the same time or just one at a time?
I just put both in first, then test individually after because I want dual channel running so it made sense to test that first. How I did it was basically this: Run at 2.8v first, seeing how high I can go passing 30-50 tests of test #5. When I pick up errors, up the voltage by point one, then test how far I can go again. All the way up to 3.3v. So the memory was being tested constantly for about a day. Maybe it had a mini burn in but certainly not a proper one. Overnight I left it to run at the highest preliminary stable clock the full test which might have helped burn in somewhat but not a proper burn in.
just got myself a new toy today..
TMSP PC3500 BH-5 1A4T
still in burning in process.
currently doing 2-2-2-5@3V 200mhz
running for almost 4hours no problem.
later will hit higher volt with same setting.
after complete my 24hours marathon will
push this ram to the limit.
Twinmos speed premium 3500
Only @ 3.1v
http://cranox.com/Rest/Overclock/32M-240-31v.jpg
Somethings wrong there, my system at 250Mhz x11 does SuperPI32M in 26m54s.
I was about to say exactly the same thing.Quote:
Originally Posted by uwackme
Hmm, i never pay attention on the time ( only see if mem is stable ) but now i also see that my time is at least 5 mins to much :(
Gonna check some things out
Hmm , even with some tighter timings i still get aprox 34 mins :(
Everything looks right , maybe i need a format or something ...
Here is shot of sandra and 1M pi
Edit : I think i know how it comes that pi 32M is so bad.
Even when i let it run in realtime the cpu usage is NOT 100% , somethimes 10% , 26% , 88% enz..
When i do 1M then almost the whole time is cpu usage 100% , weird things here
Can somebody help me :stick:
I made a new thread , help is apriciated
http://cranox.com/Rest/Overclock/sandra&1M-1024.jpg
Some TwinMOS OEM 256mb stick @ work
Click here (273mhz 2-2-2-5 benchmark action)
:toast:
Maybe you have Thermal Throttling or something on, effecting CPU % activity. Or some other thing in the background running.
2x512 TMSP PC3500 1A4T, consecutive S/N. Abit AN7 with 2T BIOS, 120mm fan @ 1000 RPM blowing over the sticks. I can hit ~232 @ 3.2V Memtest86 & 24h Prime95 stable - I keep them like this for 24/7 SETI@Home.
BUT, whenever I try to go over that, say 235, no matter what DIMM each stick is put in I get random errors in memtest #5 in the 995-1024 MB range - doesn't matter if I put stick 1 in DIMM1 and stick 2 in DIMM3, or stick 2 in DIMM1 and stick 1 in DIMM3, the same thing happens - always errors in that same 995-1024MB range. I've tried all possible combinations for DIMM1, DIMM2 and DIMM3 - nothing helps. This leads me to think that it is not the RAM's limits I've hit, but rather the mem controller's...? As a side note, I can't get the system to boot if I put them in Single Channel (DIMM1 + DIMM2). Anyone know what I can do to get rid of the same constant errors in the same range?
been letting the pc3200 1a4t's i got settle in at 3.1v's to get em ready since i was having issues when i first got em'. i'll be posting tomorrow how well letting them settle did. i see others doing pretty good with these sticks but i seem to be having problems with mine. i hope my memory controller hasn't gone bad.
btw i heard that the new bh-5's coming from twinmos don't have that 3.3v spd chip, instead it's 2.5v's now. i hope i'm saying this right. anyone know the truth? it could be the reason some people i know are getting wierd stability issues over 3.1v's.
Is that 1T or 2T?Quote:
Originally Posted by gundamit
I use one Speed Premium PC3200 and one Speed premium PC3500 with a Asus A7N8X-E and Vdimm Mode. I can hit 230MHZ with 3.2V and CPC ON! :cool:
has anyone any idea how this memory works on an intel platform ?
I would love to use some TMSP on an Abit AI7 with a booster. I get about 3.4 V with the booster.
Any help would be appreciated. The proc is 2.8e D1 ( i think ) .
It should work roughly the same. It all depends on the northbridge though. AMD's as you know have on-die memory controllers whereas Intel platforms use the chipset. That's about all the information I can offer you I'm afraid, not much of an Intel man myself. It should perform roughly the same though.
I have tried my TMSP 3200 CH-UTT in a friends Asus P4P-800 with booster, and I can´t hit 230 2-2-2 with stability, but the sticks run perfectly in my Ultra-D @ 270 2-2-2... ;)
BTW, it is only my personal experience, and in other intel system they could perform great...
nice clock with the ch-5's but the bh-5's seem to not be doing as well on average as the TMSP ch-5's. sad thing is all the new TMSP's are bh-5. i'm waiting for an rma cpu cause my mem controller was bad. individually i can get 240mhz with one and 245mhz with the other @ 3.1v's(no burn in). any higher voltage and i get errors even when i go stock 200mhz.
True... the bh5's seem to suck a bit indeed, I got to test 2 sets of 2x512 Speed Premiums BH5 PC3500 at a Lanparty a while back, and man, not even 230 stable @ 3.7v!!!
Then I popped in my UTT CH5 Speed Premiums pc3200 and.... 260 no problem (3.45v)
Today I received my second set of pc3200 CH5 so I'm very anxious to try these out... I hope they do 267 too (max on set 1) so i can run 2GB of mem :D
That'll probably be @ 2T then but well... 2Gb just....rocks ;)
I still like my MSI Neo2 very much ;) especially with the 3.3v--->Vdimm mod
(OCZ Powerstream 520 PSU with the 3.3v line @ 3.5v)
Quote:
Originally Posted by omga14
Okay, both these statements can't be true at the same time. Garrett, did you get your new TMSP sticks from newegg?Quote:
Originally Posted by Garrett
Only PC3500 SP's are BH5. I looked high and low on this forum for any updated info on this and found nothing to dispute the fact that PC3200 TMSP's are UTT-CH.
Hi XENOLITH,
I don't buy the 512's, just 256 sticks.
The last 4 SP PC3200 sticks I have received from NE are 1A4T[utt/bh5?]
Several weeks ago, the 4 SP PC3200 sticks I got from NE were AA4T[utt/ch-5?]
These are single sticks, not sets.
are these safe to run at 3.6v with a 60mm fan on them?
over at EOCF ive seen them dieing at that speed, but we know that site is for light weights :P so decided to ask here :D
I had the ch-5 version running 24/7 at 260 with 3.5v. Had 2 80mm fans blowing over them. After about 2 months, 1 died completely, the other would boot up, but only be recognized as 256. There could be some issue with the dfi board, but I would say these can't handle the high voltage over time. Rma'd them to newegg and received the bh-5 ones. Not nearly as good though. Max at 245 with 3.6v. I've pretty much given up on these. Running some tcc5 now.Quote:
Originally Posted by Korpse
next week some UTT-CH from the new charge are going to be burned on my DFI nF4 :woot:
Okaaaaay... maybe I should have told that the second set is a used set ;)
And besides, I've never ever ordered something from Newegg, the stores here in Holland are pretty much ok so I don't need Newegg... yet :D
But the second set does 260 2-2-2 with 3.6v... not really fantastic imo... the 1st set does 265 @ 3.5v so that's a bit better.
I just tried all 4 together in my MSI Neo2 board, but that was not a very good idea :D
Since it's 4 double sided dimms together you can't run at 200 or higher, 166 max.
I could not get them stable, not even at 133... ah well 2gb of ram IS a bit much eh ;)