Printable View
ok guys, hooked up my booster and an msi neo plat board, tested with 2 dimms 512 at 3.2 volts, managed to hit 253 fsb on both sticks at 2 2 2 10, for some reason I feel the board is limiting any more, it didn't read 3.3 and up volts :/ , didn't have time to memtest but I did run a few loops of 3dmark and for those few loops, it ran all the way through. I'll update when I can get a good memtest in. I hate the neo plat board btw. freaking xp-90+ dimms + booster in 4th slot is tight fit,
story of my life with the neo2 and booster woes. We need some ram that has a booster built right into the PCB. You plug a molex right into the top of the ram stick and bam! I will be demonstarting this new ram with the integrated booster on the 2005 XS tour! Call Ticketmaster before the tickets to this amazing event are sold out!Quote:
Originally Posted by fareastgq
gah, I keep missing these :D
Like i said before, jimmspc-store.fi ;)Quote:
Originally Posted by godsa
hi
do you know something about the V-Data? it's said that they have the same chips
http://usuarios.lycos.es/jopemerindo/DSCN2325.JPG
and there are very cheap (1GB = 130€ ~ 170$ )
do you think that the v-data can do something simmilar to the UTT?
regards,
well, here is my V-data PC3200 (UTT) :)Quote:
Originally Posted by steve-0
V-Data
40503
VDD8608A8A-5B
Totaly stable:
http://xs.to/xs.php?h=xs14&d=05054&f..._3.36v_nr2.bmp
Another one but it isnt 100% stable yet.
http://xs.to/xs.php?h=xs14&d=05054&f...2_1T_3.56v.bmp
Hipro5 ~ I have the same board as yours. What is your GAT setting? My TMSP are 2 X 512MB AA4T/44D/0503. I can only managed 2-5-2-2 CPC off/PAT disabled at 255 FSB @ 3.6v but not fully stable/memtest error. Pumping more juice up to 4.0v also doesn't help. Is it my board as my 2 X 256 BH-6 and 2 X 256MB BH-5 (both pair KVR) also stuck at the same FSB. TIA.
wow, i didn't expect that nice results! is that screen at 2601T at 3.36v? :eek:Quote:
Originally Posted by wez
Oh yes it is :toast:Quote:
Originally Posted by steve-0
nice... i hope to have simmilar resultsQuote:
Originally Posted by wez
regards,
Yes it is you board ONLY IF you haven't mod it yet......... ;)Quote:
Originally Posted by bwan
Upping the VAGP at 1.95Volts and Vchip at 1.52Volts I think that you could do about 263 - 268MHz..... ;)
.......and yes......with 2 x 512MB you MUST have PAT = Disable and CPC = Disable as well........Also try : 2 - 5 - 2 - 2 , 32min , 64MB , Auto , Normal , 4 (or 5) and NOT Auto , Enable(or Disable) and Disable.....Make sure to have your AGP/PCI lock at 75/37 or 76/38 and PCI Latency = 64.......
Yes...... :)Quote:
Originally Posted by esdee
anyway to disable pat on IC7?
everyone, is this "the one" (BH UTT that is)?
0448d
tmb7608f8e50I
white sticker
sn448w45m41074822
pnm2g9j16a8att9f081cadt
model m2g9j16a-tt
A guy is trying to sell me these TWinmos 2x512MB, wonder if it's the right BH UTT one? Thank you.
Just check back the thread, this is very unlikely to be BH UTT chip. But I'm still waiting for some xpert input. For now let me have some hope (although it might be shattered pretty soon :( .
No, the CADT at the end of the p/n should be AA4T
thx a lot, guess have to turn this deal down (he's selling it for like $120), too bad.
PN:m2g9j16a8att9f081cadtQuote:
Originally Posted by undeadlord
CA = 1,1u
DT = PSC chip
it is not Winbond UTT :slap:
Thanks. I'll do the mod first and post a result later.Quote:
Originally Posted by hipro5
Ouch...Quote:
Originally Posted by tictac
I guess this is why peoples love Hipro5 so much. :toast:
1.56 vchip, 1.95v vagp as suggested by him. 260 FSB but I am not finished yet. The VGA card at stock coremem.
bwan which memory ? your sig tells BH-6 and thread is about TWINMOS.
sorry that was my old rig. This is 2 X 512MB TMSP. Same week as yours. 0503/AA4T/44D M.TecQuote:
Originally Posted by TEDY
Are you saying putting 1.95 on your agp card helps overclocks? Why is this? And is 1.9 just a bit much? Will it break your card? Does this work on all motherboards? I have a DFI NF2 board and am wondering if raising my agp above 1.6 will net me any results.
Raising VAGP nothing to do with a VGA card. You have to vmod your VGPU on your card to help overclocking your VGA. Raising the VAGP basically will add more juice to AGP slot and it's helping to gain more FSB and stabilize it as I have to raise my AGP/PCI as well. Correct me if I'm wrong. Need Hipro5 to verify this.Quote:
Originally Posted by cantankerous
bwan you're getting 260 on MTEC ? wow
what vdimm ?
Yeah tedy those sticks are hit and miss one of my sticks did 240..with errors... :stick:
the other flew to 261 ...Lol..it only did 258..but when I did that step up test...it kept geting higher
Bwan ..nice job Btw
VAGP helps the overclocabillity of a system - as for an Inlel one - I haven't tested it on an AMD system to see how it does.......On EVERY Intel rig out there , uping the VAGP you always gain......Vchip helps a bit though......VAGP IS feeding the Northbridge with more juice and performs a lot better at 1:1 , 5:4 and 3:2.....Uping it at 2Volts though you might "loose" a bit of VGA's Core MHz.....Quote:
Originally Posted by bwan
I also found out that uping the Vio(from the PSU 3.3V line) at about 3.65Volts , helps overclocabillity as well...... ;)
Way to go bwan...... :toast:
3.4v :banana:
http://freeweb.siol.net/akersman/265.JPG
littledevil memtest stable ? some proof ? :)
Lol..everyone is kicking some arse with these twinmos!!! :banana4:
Show us some super # !! Litledevil...and congratz :toast:
Hmmm I have my AGP voltage at 1.6 from default 1.5. I will try 1.7 and see what I come up with. I currently have my 3.3v rail at 3.42. Are you saying raising that to 3.6 or so will also help stabilize my board and allow higher overclocks? If so I will raise the pot some more.
I wish you luck cantankerous! I have an NF2 too and waiting for your result :) I will bump my OCZ rail up too Good luck.
Thanks Negative Design. To be honest I already have a max limit of 243 on my DFI LPB. I was hoping for more which is why I am going to try the above mentioned. I am going NF4 soon anyhow so I hope to get at least 260 on that setup. I just changed some cooling in my case and bought a Zalman FB123 bracket to help cool the dimms. I hope for greater results. I will surely keep you posted.
Do you have VX/UTT on your NF2? If so what are you getting from it?
Quote:
Originally Posted by cantankerous
Im not doing to good on mine :( I have BH5 and I'm stuck at 240FSB I think I need more volts because I can only stabilize it by relaxing timings :(
Are your sticks getting hot? alot of ppl said their UTTs stay cool,my BH5 on the other hand is HOT! 60C lol.
My BH-5 doesn't even get warm dude. I'm doing 10x294 2-2-2-6 @ 3.5V
Do you have a screeny of that?!!!Quote:
Originally Posted by Klisp
That's amazing :toast:
they don't get warm @ 3.5v at that speed? lol gotta be kidding me! :p:
yea mine are pretty hot. The max I seen them get to is 51c at 3.3v but that was in an enclosed humid room. Normally under memtest they get to about 48c and under gaming about 41c.
screenshot please ;)Quote:
Originally Posted by Klisp
Man I can't wait till get my dam sticks..is it monday yet?
Everyone needs to post some pics if they can
Mine do :) I put a temp probe on the bastards and it gave me 53 load.Its strange but all 3 running 2gether get HOT! especially on memtest at 230FSBQuote:
Originally Posted by Klisp
But when I mix my mushkin with my ultras they do 243FSB 1T and stay warm! I guess bad combo/sticks get warmer then good ones,I just found this out last nite btw,after mixing my 5 sticks.
stable?Quote:
Originally Posted by Klisp
ive never even heard of that with 2x512.
edit - oh wait, maybe 2t will do that...
anybody knows what kind of ram is used in PCStats review? http://www.pcstats.com/articleview.cfm?articleID=1724
I saw the "dimples" in one of the shot, but not sure if they are AA4T's or not....the SpeedPremiums won't post at CL2 :(
yeah thats them... -44d with dimples :)
only fed them 2.8v hah, if only they knew.
I think they knew...but didnt say! bcuz they dont want ur hungry geeks to want them more!
Or maybe they didnt want their sponsor OCZ to get mad who knows.
ok, thanks guys. i thought they were aa4t's :(...
They are aa4t on the white stickerQuote:
ok, thanks guys. i thought they were aa4t's ...
The 44d refers to the last few numbers on the memory chips itself.
Hope I helped
*********************update*********************
Recieved 4 sticks today all aa4t...
Stuck one stick in..has been around 40 minutes..still going strong at
3.3 250 fsb...Going to jump it up and try 260..
2 2 2 2 2 11 13 15 Timings.might not be the correct ending order but you get the jist of it.
I'll be doing the 3.1 to 3.3 test as well ..I'll contribute my findings to that thread
Anyone know why I can pass memtest 260mhz 20 loops of #5 and I can't get into windows?
Give it more volts :banana:Quote:
Originally Posted by mcbarnet007
that's not the solution....it is already at 3.5v
if it pass memtest it should load windows right? considering I have 4000+ running a 9x multi.
Is memtest screwed up? or windows and memtest uses different setting?
loading windows is usually harder on the system as true dual channel comes into effect not to mention the loading of drivers etc. Memtest is good to be sure your memory is stable but cpu, cache and mem controller are totally different things that come into play during the loading of an OS. You can either raise voltage like mentioned above to give the memory a bit more juice when loading the OS or back of a few mhz at the current voltage.
Try raising your vcore, maybe its the proc and not the ram.
4000+ uses clawhammer, which has CG mem controller(prob the best one right now). At 9x 260mhz I am at 2340mhz which is below stock for 4000+. Even when I use 1 stick (single channel) I have problem with window based memtest at 255. I can run memtest at 260mhz 3.33v but can't even run 3dmark01 at 250mhz 3.5v.
Does VX have the same problem or just twinmos SP?
Good idea, or better yet lower muti far below your CPUs limits for more RAM FSB headroom.
I have the same problem,I'm memtest stable at 250 but windows stable at 243 your sticks may be faulty never know.
lol, but isn't memtest suppose to test to see if your ram is faulty or not?
yes,but like cantankerous said, it tests only ram :) not ur controller/cpu or any other factor(that I know of) that affects ram overclocks.
ok, I just ran 25 loops of memtest #5 at 5x 260 2-2-2-8 at 3.4v on DFI SLI-D with all the other timing set to auto. Not one error. Boot into windows and try to run 3dmark01 and it crashed on the second game. Should I up the chipset voltage or the LDT voltage?
Yes I know, but I know the CPU/mem controller is good, test it on TCCP and runs 275mhz 2.5-3-3-7 at 2750mhz just fine.
How is it going now?
What type of error did it give did you check?
Memory is being cooled?
Maybe you video card can't handle those timing being throw at it? Is that possible?
Thought I read something about that on this forum.
i follow the instruction of giving it more voltage. So I throw 4.0v at it and now it is completely stable at 265mhz in windows. 3dmark and everything. I just hope it doesn't die. someone really need to come up with water cooling block for memory, even 1/4 tubing is ok, just make it a seperate loop.
4 volts may be getting a bit hot, better cool those suckers to be safe. they aren't hot when your just surfing the web or writing a paper, but if you start playing games or somthing, they'll start getting warm > warmer > hot
haha, I have fat ass copper ram sink on them and a 92mm fan blowing at it. Even memtesting the ramsink is kinda cool. But one of them died.....
Hmm...funny. I plug in my back up stick and now I can run at 260mhz and 3dmark01 no problem. I bet you anything one stick killed the other one cuz they don't get alone. Oh well survival of the fittest.
Quote:
Originally Posted by mcbarnet007
Are you suggesting maybe something like this?
http://www.kaltmacher.de/viewtopic.p...er=asc&start=0
im gettim some speed premium AA4T soon :WOOT:
hello, one shop give me this number of the twinmos memory:
P/N: M2G9J16A8ATT9F081AA4T
MODEL: M2G9J16A - TT
the chips are winbond UTT???????
yes, they are.
yeahhhh :banana4: :banana4:, in spain is very difficult find a twinmos memory, thanks
no problem. go and buy the sticks as soon as you can. and as many as you can afford. ;)
tell us what the top 5 letters/numbers are on the chips themselves. should be like 04414, then, on the 2nd line what the last 3 things are. should be something like 44D, or 50D.
price? :P
150€=191.76 $
not bad for europe. only about $15 (~11€) over our prices.
spain in technology is very bad for buy components, today is arrive the dfi nf4 sli-d, sli-dr and ultra-d, the price of ultra is 162€ around 200$
http://www.newegg.com/app/ViewProduc...218-062&depa=0
Just received this dual pack. AA4T, 50D. production date is 05034 (translate please). Looks to be a better value than the SP.
its winbond utt...
made in 2005, 3rd week. 50D... "supposedly" slightly worse then the 44D, but i dont really know how true that is.
hmm not everyone seems to be too excited anymore. the high must be over, lol. I guess they rate it a little lower due to the fact that it is a dual pack. Whatevers. Only volts will tell.
swuacheneger geeze your english sux :)
jk/ :P
:stick: hehehe
there really hasn't been enought testing and research done on those sticks yet..Quote:
its winbond utt...
made in 2005, 3rd week. 50D... "supposedly" slightly worse then the 44D, but i dont really know how true that is.
But I did manage to get one of those sticks to 261 ..no errors..
Another stick 240..had a couple of errors
That is where the big drawback is..one flies..the other died...LOL
With the speed prem..most people get 250 fsb each stick..its sort of steady
150€? not bad at all... i payed 170€ for mine UTT in Portugal... :mad:
but hey... gotta be positive about it, atleast i live in a country with beatiful sunshine, beach's and women. :D
@ godsa , are u sure about that?
http://www.newegg.com/app/ViewProduc...218-062&depa=0Quote:
sure about that?
Sure about what I said? I said alot...the memory above...only a several people have.
Magority of us have speed prem
Im refering to the memory linked in this post ..
So yeah Im sure about the testing..less testing and research on those..because less people have them
Am I sure about 261 ? Yeah Im positive.
The only one really geting lower then 250..was tedy and the starter of this thread..that 2 people out of alot..well and u according to the vx info thread
And as they said probally their boards.
Here is all the data collected so far..check for 50d and that memory
http://www.xtremesystems.org/forums/...ad.php?t=50699
check it..
That thread should really distinguish regular twinmos and speed prem...
Would give better data analyis
hahahahahahaQuote:
Originally Posted by godsa
it should also distinguish between nF2, nF3, nF4... socket 462, 754, 939.... 939 newcastle, 939 winchestor... brands of UTT.... and oonnn and oonnnn and ooonnnn....
short answer... unless you manage to slow down the rotation of planet earth, giving me more hours in the day... it jus aint on the cards hehehe
Thanks for the recommendation anyway dude, but I jus cant be bothered... there I said it :D
Quote:
Originally Posted by godsa
i just got these today. so far memtest stable at 262 2x512 dual in dfi nf4.
they are 50D and week 0503 very nice i must say!
http://www.xtremesystems.org/forums/...chmentid=24300
nice trans am, wot vDimm are you giving them?
and are you using 5v or 3.3v jumper?
Nice FSB there trans-am...are you selling them ? :slobber:
Great job man!! go for 270! :banana4:
What volts?
Speed Prem right?
ok lets put 4 x 512MB on that DFI NF4 next... :hehe:
@ godsa , i was asking if u were sure about that SP vs regular TwinMOS results, as most of SP have same chips (tho 44D are SP exclusive.. but i haven't seen any advantage on them yet) and pcb's (the BP is an exclusive for SP aswell.. yet to be seen any advantage) than regular ones... so, it's basically random luck i guess.
I was tryn to burn in my Twin Pack non-sp last night and i realized heat is an issue with these UTTs. W/o a fan, i get 100 something errors at 220fsb at 3.2v. With a small fan on it, it get on average about 5 errors per pass over 400 passes for 6 and half hours.
I lowered it to 214 and Im memtest clean but I was not able to boot into windows. I have this theory where if I use a fan during memtest, the results will be schewed. Testing the mem without a fan will give me a truer memtest result.
Gotcha...Quote:
godsa , i was asking if u were sure about that SP vs regular TwinMOS results, as most of SP have same chips (tho 44D are SP exclusive.. but i haven't seen any advantage on them yet) and pcb's (the BP is an exclusive for SP aswell.. yet to be seen any advantage) than regular ones... so, it's basically random luck i guess.
As of the research now im pretty sure...if we get more people with reg twinmos to test..we will c...
I bought reg twinmos it did 261 error free...the second stick did 240 with couple of errors..
Then I bought 4 sticks of prem...all 4 sticks will do 250 fsb at 3.3 on my board ...One stick will do 255 error free another will do 254 error free...
After burning them in..I can do 255 on both in dual channel ..because Im too lazy to burn them seperate...No Errors
Today Im burning them in at 256...6 am till 7pm..then Ill let them "CURE" and check them in the morning at 256 and see if they pass
so can you explain teh burn in process a little to me?
I have normal Twinmos non Speed premium Mtecs and I have Nrain power PCB. So BP is NOT Sp only.Quote:
Originally Posted by enok
Quote:
Originally Posted by trans am
How many volts for 262 trans-am?
nice chips!
@ Entity Razer , when have u bought them?
either way it's less (way less) common to see non SP with BP pcb... congrats.
@ Trans Am , wanna sponsor me? :p:
this was 3.3v. I am going to try some higher volts later. This is with the dfi jumper rocking off the 5. I think with the right settings and volts I could get them up to 265 memtest stable. I tried 265 with my 262 settings and got 30 errors on the 1st pass of test 5. So I think some vdimm and vcore would help.
The part that gets me is that these are just plain twinmos on the plain jane 6 layer printed circuit board. This is not the Brainpower. And it's killing my VX at the same settings in dual channel. According to the dual results, if I put these in my nf3 754 dfi board, they could easily do 270. I am basing this on my results with the VX in single(264mhz) as opposed to my results on dual channel(255mhz) with the same ram.
well guys, after my buddy played games for a weekend with a stick of my SP, my booster broke :( the Neo won't boot up with it in anymore, just beeps, and that was only at 3.4 volts :( I don't see anything wrong with the booster, it just won't work... Anyway, the fastest the stick I was burning in went at 3.2 volts is 250 fsb at 2 2 2, I was trying to burn the stick in....blah, Anyway, they aren't better than my corsair xms pro (which run 265 at 2 2 2 a stick @3.4 volts) and I like the LED's, they allow me to see if a board is toasted or not, so I guess I really have no use for the twinmos. Granted they are cheaper so it's best bang for the buck atm. Off to ebay they go tonight, as I don't wanna pay a restocking fee at the egg... and i can't really rma them becuase there's nothing wrong with them...blah. Between that and rma'ing my booster, I've had kinda crap luck with burn in atm. Bad thing is, I never really got to see what the 2 sticks I didn't burn in yet could do at 3.6 volts. Stupid Neo and it's ovp... I can't wait for DFI nf4 AGP.....
262Mhz with 3.3v? WOW :o i want those! :toast:
i just cant believe it ...262 with only 3.3v
well you never know. They might just be lucky sticks. Luck of the draw.
trans am, it would be nice if u could tell us how they scaled with diferent voltage increase. :)