what i noticed with hypers you cant have nb set too high or you gonna have many errors in memtestQuote:
Originally Posted by CryptiK
i still find it nuts likes cas7 better than cas8
900 cas7 i can pass 900 cas8 fails it memtest:ROTF:
Printable View
what i noticed with hypers you cant have nb set too high or you gonna have many errors in memtestQuote:
Originally Posted by CryptiK
i still find it nuts likes cas7 better than cas8
900 cas7 i can pass 900 cas8 fails it memtest:ROTF:
I did find for absolute stability, vNB needed to be adjusted very delicately.
Also the specific kit you get really defines how they will scale on CAS 6/7/8 etc. tRCD is more clock limiting than CL for Hypers though. My kit, as with most Hypers, scales well with CAS 7, and CAS 8 gains me around 100MHz more at the same voltage using X48, around 200 MHz more on X58. Some kits (the lowest end cheapest Hypers) seem to be more picky from what I've seen, and they do not clock as well as the better kits either, which leads me to ask, what Kit do you have?
DDR3-2000 CAS 7 sticks would be the optimal kit, but if you cant afford it, the DDR3-2000 MHz, DDR3-2133 CAS 8 or 1866 CAS 7 sticks are really what you want to be getting. BBSE performs very closely to low end Hypers. Remember, X58 clocks this ram better at the same given voltage.
Here's a couple of my benches using my Corsair 1866 C7 GT's.
SPI32M @3.94 GHz / 563 FSB / tRD 8 @ 1.47vNB / 900 MHz / 6-6-4-20 1T tRFC 60 @ 1.76vdimm
http://i720.photobucket.com/albums/w...v900MHz6-6.jpg
SPI32M @3997 MHz / 571 FSB / tRD 8 @ 1.47vNB / 913.6 MHz / 6-7-4-20 1T tRFC 60 @ 1.76vdimm -- So close to sub 11min @ < 4GHz!
http://i720.photobucket.com/albums/w...7v900MHz6-.jpg
SPI32M @ 4.25GHz / 500 FSB / PL8 pullins 1+2 / 1000 MHz / 7-8-7-24 1T tRFC 72 @ 1.72v
http://i720.photobucket.com/albums/w...ns121000MH.jpg
i got the patriot sector5 2k cas8
just got my p55-udrp+860 will see how they do today :yepp:
I kinda squeezed this one out, dont mention the time please, it's untweaked and on windows 7.
http://img.photobucket.com/albums/v6...trd8600fsb.jpg
The purpose was to finish 32M at 600 FSB with PL8.
Thanks to the crappy heatpipeassembly it took quite some volts to the NB as you can see.
I even had a delta fan pointing at the hpa that i could still hear upstairs in the bathroom, lol.
One very nice gentleman is going to send me a pair of great Cellshocks so that i can really start testing the board. :)
after a while, im back to o/c land! :D nice results guys, good work :up:
Unfortunatelly, my E8600 died on "battle", so im waiting a replacement, to do some progress.
These are my results with my Q9650:
http://img195.imageshack.us/img195/1915/19411566.jpg
Still some memory tweaking needed :)
well 2gb hypers didnt do well i bet they were low bin
i got me some 1gb stix of hypers commin in :D i thonk that should be fun
if i can get 2000 cas 8 i be happy camper if cas 7:shocked:
i havet been stable @ 2000 with any stick yet lol
hello nic do you post your timings ram of bios?Quote:
Originally Posted by N.I.C
Gullo, from the 40 posts you've made, surely 38 must be this very same question?
Why don't you take some time and test yourself with memtest86?
Everything you need to know is at memset on my sceenshot.Quote:
Originally Posted by Gullo92
i have find stability with my q9550 455x8.5 with my corsair dominator 1820 7 7 7 18 4 60 10 5 8 5 6 4 7 4 7 21 7 1 8 8, i would test the timing of nic.Quote:
Originally Posted by Zeus
Today i finished dealing with the damn heatpipeassembly.:mad:
I've put an Enzotech CNB-S1L on the SB, shame it's not sitting square but i'll do till i find a good alternative.
For the mosfets surrounding the CPU area i got two Enzotech MST-81 sinks that came in with some bend pins. :( Fortunately i could bend the pins back and it all looks ok now.
The NB got an EK NB S-max block and i was trying to put some small Enzotech sinks on the mosfets but the adhesive tape was too weak and the sinks kept falling off. :mad:
I ended up going to the local computer shop getting some Artic Silver adhesive and glueing the little buggers in place.
I wonder how you guys managed to get 'm to stick. :confused:
Anyhow, NB temp went from 41ºC down to 31ºC with 1.4V, so with 10ºC off definately worth it.
The stupid SB is still 43ºC but i don't care much about that, NB temp is much more important for high FSB. :)
Here's how it looks, sorry for the bad quality pics, i took them in a rush:
very nice Zeus clean and simple, what ever happened to your BP sb/nb block ?
same here even with the bigger ones they will not frickin stick :)Quote:
Originally Posted by Zeus
:up: very impressive.Quote:
Originally Posted by Zeus
Good work Zeus:up:. Try find out some Ceramique adhesive pads, these are the sticky ones ;).
Thanks guys.
The small sinks are pretty secure now with the AS adhesive. :up:
Just ran some SuperPi and the NB went from 32ºC to 33ºC and now it's stuck at 33ºC. Still a nice improvement, especially the load temp is very good now.
The stupid SB went up to 47ºC after i closed the case so here's something i could improve on.
Quattro, i sold the Bitspower together with my old RE, the RE i'm on now is a new one.
Ater going AMD i just wanted to get back on the RE, i absolutely love this platform although the Crosshair III Formula i had was a very nice and beautifull board.
The olny thing that's keeping me from tearing things up is the RAM but that's taken care of by a good friend. :)
I don't have friends like you it seems :p... good luck man !!
Hehe, yeah, he really made my day. :) Coincidently he's also from Belgium and he's also on this forum. ;)
Expat Griz ?
Some one rang?!?!?!?!?!?:rofl::ROTF::rofl:Quote:
Originally Posted by Leeghoofd
Yup, still have another 2X1 set here...unfortunetly I need a good 8600...this ones a dog:shakes::shakes: topped out @ 5.8:shakes:
that suxors man , mine did 6.1Ghz 3D stable (till now) 6.3Ghz pi stable and is being tortured by Massman for the F1OC FSB competition (hope the CPU will come out good)
Ok fellas, been a long nite here.
Got 2 goals on my mind, nr 1 being breaking 8 seconds SuperPi 1M and nr 2 being 32M under 9 minutes.
About a year ago i did a 9 min 20 seconds 32M run at 4840 MHz or something but back then the OS was tweaked very well it seems cause now i got the same time at 5GHz!! :mad:
Anyways, it was nice to see the cpu rip 1M at awefully close to 5.1GHz and completing 32M at a staggering 5GHz.
With some more cold i will get better results i'm sure.
Did not achieve one of my goals yet but i can say i came pretty close.
I'll leave you guys with the screens for now, more testing to be done when the Cellshocks come in. :)
http://img.photobucket.com/albums/v6...ctory/9062.jpg
5GHz 32M
http://img.photobucket.com/albums/v6.../32M920578.jpg
The time still sucks for the MHz, seems i lost my touch. :(
Oh btw, all done on watercooling with a fresh 12ºC evening temp. ;)
Zeus, very nice on the cooling and benching. do you happen to have a temp comparison with the heatsink assembly on with the NB block on top of everything vs the direct contact of EK nb S-max? right now, i have the stock assembly on with a swiftech mcw-nbmax had my temps are around 39C i think. on air, it was 55C. looks like a direct contact with the NB shaves off another 5C compared to the NB block on top of the stock sink assmebly? and don't worry about the SB temps... it does absolutely nothing for your OCs' :lol:
1.53v is not that high yet. pump a few more volts in there and up the fsb to break the 8sec spi 1m? i wanna see you do this on water also :up:
Great work Zeus, the new sinks & block on the board looks great and the NB temp is very nice. I've been running a nice Q9650 on my REX recently and the NB temp
hits 54*C @ 1.43v cooled by 80mm fan while running Prime large FFT. It needs water, the stock HPA is no good.
Congrats on 1M & 32M runs, nicely done. What vNB for that FSB 32M? PL 9?
Thanks guys.
It really is a pleasure pushing this board, once it's on the edge it just errors out in SuperPi instead of bluescreening or rebooting like other set ups do.
About the temps, i have only had it running with a Swiftech NB max block and MX2 tim plus plastic washers added and temps idled around 41ºC with just 1.4V running my daily 10x450 pl6
Once i started HCI memtest 48ºC was easily reached within minutes.
Now with the block straight onto the NB, idle temp is around 32ºC and load is about 33-34ºC! :)
The only thing that needs some work is the SB. Due to the lack of airflow through my case it reaches 47ºC idle, what used to be 43ºC with the hpa.
Cryptic, i think i set 1.7Vnb at the sart of the benches, just to be safe and because the NB runs fairly cool now.
It will 620 PL 9 with ~1.63Vnb.
Once the winter rolls on sub 9 seconds 1M will be done, i'm sure. Another 50MHz should be enough. :)
Sweet clocks! Well done Zeus m8! :clap:
Hi Joebar! Thanks and good to see you around again. :)
Where can I get window7 drivers for this board?
I might be wrong, but I was under the impression you use the appropriate Vista drivers; 64 bit W7 = 64bit Vista driver etc.
I'm sure someone will soon correct me if that is not right.
you are absolutely right :up:Quote:
Originally Posted by CryptiK
:yepp::yepp:Quote:
Originally Posted by rurik
Hi I understand 2x1 or 4x1 kits with GTR/GTS chips are optimal for the RE
I currently have a 2x1GB Team Xtreem that I believe is a JNL, clon did manage to get a GTR one though ()
So which kits should I get if i wanna upgrade, either 2x1 or 4x1.
I been searching high and low for Cell Shock and CSX, the latter I believe is not available in the US?
Mushkin is another choice?
Is this a good one?
http://www.memoryc.com/computermemor...800cl7kit.html
Will cost me 100 Euros to get it to my country :(
You'll be hard pressed finding any decent Micron D9GTR based rams from my own experience, which could be read in this thread: http://www.xtremesystems.org/forums/...d.php?t=235807
Your best bet is ebay, forums for sale sections and the like.
Since D9GTR/GTS is obsolete for a while now there will barely be any ram modules with thise chips left in the stores.
If you're desepeate for a kit (like me) you can always get these: http://www.hpm-computer.de/index.php...DR3-2000-.html
I recieved my kit two days ago and have to admit that they're not as good as the CSX Diablo's i bought a year ago.
This kit doesn't like Trfc set to 48 and need a minimum Trfc of 60 to even boot. Also, one stick does not like Trp set to 5 and refuses to boot with Trp 5.
Furthermore, i tried pushing for DDR1920 last night but couldn't get it, not even with 2.21V. :( Maybe i've overlooked something and have to try more but i think the old Diablo sticks were better.
Getting some used 1 year old modules would be your best bet but you're risking buying "worn out" sticks.
Cell Shock is out I guess. Their GTR models are discontinued. CSX are way too exp....
So I take it that Mushkin is my only choice among those 3 brands?
The Cellshocks you linked are D9JNL based, stay away from them unless DDR1800 7-7-7 is good enough for you.
Dunno 'bout the Mushy's, hard to tell if they're GTR/GTS.
Cellshock black 8-7-6 and CSX Diablo are for sure.
Given the link posted by someone in this thread, there's only 3 kits with GTR chips, of which 1 is discontinued.Quote:
Originally Posted by Zeus
There's only 2 kits left :(
Cell shock I think they are discontinued whereas the CSX are very very exp
Hmm?
True, hence my sentence in post #2531 you'll be hard pressed finding a kit.
Try to find some used modules, older Gskill HZ are either GTS or GTR (newer are Samsung) and clock pretty well and so are some revisions of Corsair Dominator.
Google for ddr3 ramlist and you will find a list of ram that contains what chips.
I found that there's quite some difference between GTR chips, some like tight Trfrc and Trp where others don't so good binned chips are a must.
Another option is to go with some Elpida Hyper based ram although it won't run timings as tight as GTR will.
Zeus mate, which is the recommeded bios for high FSB on this mobo ?
Having a daft time stabilising 620FSB (could be purely mobo limitation ofcourse) for 3D benching... she goes up nicely to 610FSB no biggie, but for 620 I have to pump 1.7Nb voltage , while for 610FSB I only need 1.57 lol...
had a hard time to be stable after 610 even with 1.8v lolQuote:
Originally Posted by Leeghoofd
You use what cooling of northern bridge?:eek:Quote:
Originally Posted by zsamz_
i moded a katana3 :DQuote:
Originally Posted by rurik
my nb never even it 40c even with 1.8v lol
with stock cooler i was 50c+ @ idle 1.6v
Hi Emptyhead (lol, forgive the translation joke, couldn't resist)Quote:
Originally Posted by Leeghoofd
I use bios 0403 95% of the time, that's the one i achieve good FSB on but i could do practically the same with bios 1202. Haven't tried much other bios' apart from 0501, that one did 600+ easy as well.
Having Performance Level at 9 might be limiting you but PL 10 is way too loose for me, important is having ai clock twister set to "light", at least for me.
If your board is a real champ you could try PL8 but that requires loads of voltage.
I can do 625 PL 9 at 1.7V set in bios (so ~1.67V real) 32M stable.
This is with the NB watercooled. It's pretty important to have good cooling on the North for high fsb.
The one setting helping to align the data bus between NB and CPU is NB clock skew, setting it somewhere between delay 100-300ps might be of great help.
Good luck trying.
Thx, already tried all that :p will try to hook up the waterblock and indeed PL 10 gives a huge dip in bandwith... Can you touch the CPU skews ? I can't with bios 1202, only NB, otherwise she won't post....
I would adjust either NB or CPU clock skew, no need to adjust both as the goal is to align the data bus between both.
Working from two sides won't make this job easier so i prefer to only use either one of those two and have had best results using NB clock skews so therefor i never really touched cpu clock skew.
Having NB temps below 52ºC is highly preferable for stability.
i got a set of 2x1 hypers and a set of dominators 2x1 1800cas7
funny i cant do 2000mhz lol
i can do 1900 cas7 1700 cas6 but 2k :rofl: is out of reach
and tips for a b00b lol
Try Dram ref voltage for both channel A & B +12,5mV.
I got the board screaming now at 600FSB with 5:8 divider (DDR1920) 7-6-5-18 Tref 48 timings 32M stable!
I tried to stabilize these settings and couldn't get memtest86 test 5 without errors, not even with 2.21V.
Then i tried 2.17V and Dram ref volts +12,5mV on both channels and voila, stable as a rock!
i had em @ +37.5:confused: maybe too much lolQuote:
Originally Posted by Zeus
i be testin again:D
I'm very close to 620 FSB stable, only get about 8 errors after 3 passes test 5, test 7 is error free for 3 passes... it ain't the rams as I can loosen or tighten timings ( currently c8-8-8-22 1T PL9 ), purely NB and ref volt related. Rams were already at +12.5, NB is at +30... I try to increase the cooling on the NB
Test 5 is very likely your ram, try loosen Trfc.
NB unstable is errors in test 2,3 and 4.
Sometimes tightning up AI clock twister helps even though it sounds odd, try moderate instead of light. (it migth lock up instantly though)
How much volts you're feeding that NB and how hot does it get?
Here's 600FSB, PL 8 and some nice ram frquency and timings --32M proof:
http://img.photobucket.com/albums/v6...8untweaked.jpg
Run is completely untweaked, just stabilty testing.
That's wehre my issue is ( this setting is just for 3D action) at 600FSB PL9 (MODERATE) I only need 1.55 volts on the Nb, for 620FSB PL9 (LIGHT) I'm close to 1.7Nb, temps at idle (last setting) are 45°C according the bios... Thx for the help man, appreciated...
Did you mess with different straps? Straps can make all the difference, don't just tie yourself to just one strap, changing straps while changing FSB.
Run 620+FSB or die trying!! :D
Maybe a silly question but are you sure your CPU can do 620 FSB+? Not all can from my experience. Not all boards are created equal either.
Zeus - nice, although 600 PL8 takes a heap of voltage, no way that would be happening on stock cooling. 32M/3D at 600 @ PL9 moderate takes 1.43vNB BIOS on my board, but I have not tried 600 PL8 as I only have 2 x 2GB ram kits to use which makes things harder than they would otherwise be.
Good luck leeghoofd.
Well, actually i ran 600FSB PL8 on the stock cooling with MX-2 tim and added washers on the back of the board, Swiftech NB max block on top of the hpa.
It took 1.73V (1.76V in bios) to run 32M but it finished. :)
As you can see my NB runs very cool with this EK block on it, 32ºC is right after i finished the run and with 1.73V in bios.
Maybe i can do with lower NB volt but since the NB runs fairly cool i wasn't too worried to push some volts.
You may actually be onto something, not every cpu likes high FSB, pushing Vpll some will raise the limit some.
CPU did 696FSB under LN2, 650FSB stable 8M under air with the DFI P45 mobo... So I'm hoping to prep it for 6.2-6.3 3D stable for the 01 runs... you got a very nice board there Cryptik... wished I had Petri's REX, that one does 700FSB 3D stable lol... he just needs a CPU that can do it :p
Ok so it's not your cpu by the sounds of it heh. I am yet to test my board with some nice 2 x 1GB GTR sticks, I have only run 2 x 2GB sticks in it, not sure how that affects things.
700 FSB 3d stable on a REX is insane, DICE/LN2 on the NB right?
This Rampage Extreme impresses me more by the day:
http://img.photobucket.com/albums/v6...y/sub10min.jpg
Not bad huh? :)
Very nice mate :) good board ya having there...
Thanks buddy, it surely looks a nice one.
I even have a feeling it's not maxed out either, the errors i got were due to maxmem/copy waza combination rather than hardware related.
Awesome run Zeus, that board sure is a nice one. So is your CPU. What vNB for that, 1.8v? ;)
Thanks. ;)
At first i was doing 630 with 1.76V set in bios so 1.73V real and it was stable.
After that i tried 632 because i wanted to go under 10 minutes (Low Clock Challenge) and got some errors so raised Vnb to 1.8V in bios and still got errors now and then.
After some more testing it dawned on me it was caused by the maxmem setting combined with Copy waza.
Still left it at 1.8V, which resulted at 1.78V in windows approximately.
With the EK NB S-max block temps never went over 34ºC though.
I thought so. 34*C at 1.8v - damn good block.
Had a blast tonite achieving both my goals with the help of some fresh dutch air.:)
First of, 1M below 9 seconds without the use of a non 45nm supporting bios:
http://img.photobucket.com/albums/v6...ctory/8937.jpg
I tried bios 0121 but for some strange reason i was unable to set the muliplier. :shrug: Maybe the bios was somewhat corrupted, i will look into that after i've had some sleep. :)
Next up was breaking into the 8 minute territory for 32M, managed to do that although i had hoped i could do it on slightly lower clocks.
Still got there eventually:
http://img.photobucket.com/albums/v6...ry/32Msub9.jpg
Pretty happy with those results, the OS needs some more tweaking but the CPU clocks pretty damn well, so does the board and the ram. :)
Off to bed now. :)
this board was just made for d9gtr 1gb stix
it boots faster and even more responsive simply flawless
i got me another set of kingston hyperx with hypers to run 4x1:D
Hey Guys, i recently purchased one of these fantastic motherboards and i like to say that i am enjoying every minute of overclocking. And i'd like to thank you all here for your wealth of knowledge that has helped me so far.
I just have a few questions about nb voltage. I see that you guys are pushing well into 1.7 Volts and i am curious how safe that is for 24/7 usage? I am at 1.65 Volts at 495FSB with a Q9650 and i am chasing 500FSB and beyond. I have watercooling with EK nb S-Max with load temps around 38C (on 32C day in Melbourne AUS, no A/C) so temps are no probs.
Also i would like to know about gtl refs? My setup seems to like + figures and i am currently at +100 - +40 - +100 - +40 which has allowed me a Vcore of 1.39 (bios) for 4.45Ghz. Are there any danger in running such high gtl values? Any info would help, thanks and keep up the good work.
P.S. Kingston HyperX rocks on this board
1980 9-9-9-20 at 2.0 Volts
It's hard to tell what safe voltages are but as soon as the value in bios changes to yellow it might be too high for 24/7 usage. But since you have good NB cooling i think you could run 1.7V as a max.(don't quote me on this if your board dies ;))
Mind you this board isn't the best for quadcores, the Gigabyte EP45UD3 seems king for high FSB on quads.
There are no dangers involved in running those GTL settings, it's a voltage adjustment not a voltage level.
Thsoe are pretty high NB voltages you need to push for sub 500FSB, my board needs around 1.55-1.57 for 500FSB with the 9550 (ES). Have you tested some different ref voltages ? I needed to push VTT pretty high though ( around 1.45)
I was testing it for the country Hwbot competition, CPU is hardlocking (also on P45) at 510FSB lol... I was ready to put it under the phase when Massman told me no ES only retail ( what a letdown ), no scores from me then...
1.65 for 24.7 is fine as long as you keep your nb cool my nb never gets over 38c;)Quote:
Originally Posted by RasCas
rex can do nicely with a q9650 but only with 1gb sticks
set gtls to what is stable for you i had mine @ 70-50-50-50 range
if you dont plan more memory stick with the REX but if you do you gonna need a ud3X
most i got was 500fsb stable and only with 2x1 sticks
Thanks for the quick replies fellas, i know about the 1gb stix with this board never got 2gb stix stable at more than 425 FSB stable and just gave up in the end.
I think i can get my nb voltages lower, but in order to do that i have to back of from a strong trd 7. But i am getting quite good memory performance, i dont really want to do that.
http://img10.imageshack.us/i/memh.png/
I'd rather tighten those timings and loose that ai clock twister to moderate and timings to 7-7-7 or 8-8-8, 9-9-9 is very loose.
At 425FSB PL 7 is not that tight, i'm running 450 FSB with PL 6 and Ai clock twister at moderate, i only need 1.39Vnb. This is with a dualcore, i can imagine a quad requires more Vnb.
edit: Just saw you're running 495FSB, still set ai clock twister to moderate and tighten mem timings so you can lower Vnb :)
Ill give that a shot, going from 490 to 495 required a jump from 1.6V to 1.65V so 500 would require more than what im willing to give.Quote:
Originally Posted by Zeus
Im waiting for the weather to cool down till i test any further. Im in a house with no A/C and we have our thrid day in a row of approx 35C and expecting another 3 days of this weather.
Will keep posted
Well, at least you got some sunshine, over here it's cold and bleach. The flue got a grip on me for far too long now and i'm spending more money on tissue's than on hardware atm. :D
Good luck testing, do you know what chips are on your Hyper-X's?
If they're Micron D9GTR you could tighten those timing a whole lot, if they're Elpida you could still tighten them down by quite a margin.
I`m stable @ 480 with vNB 1.57 real,1.6@ bios.Quote:
Originally Posted by RasCas
gtl`s is very important thing for stability... one step at left or right - and system crashed :)
Now that this board is fairly mature, there must be others that have encountered the same problem:
Intermittent sound distortion!
I'm using the fiber optical output ( rear ) and I will get intermittent periods of distorted sound from games, music tracks...anything. I've had the board replaced as well as the card but the problem re-occurs.
Window 7 RC1 with native drivers only. No enhancements.
Anyone else?
:confused::confused::confused::confused::confused:
hi i have a question, will this corsair 2x2GB DDR3 1600 1.65v built for i5 p55 or Gskill trident ddr3 2x2gb 1.65 , work on Rampage extreme ?
and look at the back of the corsair mems , rampage extreme is visible
http://www.overclockersclub.com/vima...air_cas8/1.jpg
http://www.overclockersclub.com/vima...air_cas8/5.jpg
http://i53.photobucket.com/albums/g6...ddr31600-1.jpg
http://i53.photobucket.com/albums/g6...316992gbx2.jpg
you wanna waste $$ try em lolQuote:
Originally Posted by Nurse_BC
you not gonna get nowhere with low binned crap
Quote:
you wanna waste $$ try em lol
you not gonna get nowhere with low binned crap
What? whats your point
buy em prove me wrong:rofl:Quote:
Originally Posted by Nurse_BC
Greetings! Has noticed strange feature: now I try to overclock the processor further, naturally, at increase FSB I increase NBvcore. But constantly errors in Prime and Linx.yesterday it is simple so has reduced voltage on northern bridge and OMG - the system is stable on 485 Mhz.я I think that 490 I will make too.Now voltage on northern 1.535 according to monitoring.
Has replaced standard cooling on mobo. Northern - Xigmatek Porter - the temperature is no more 53.PwM - no more than 55 degrees (it is bad?), the southern bridge-47 it is maximum.
Maybe he just wants to run them at stock speeds Zsam. I'm using the 4gb Corsair Dominators PC12800C8 on the REX Nurce_BC... there's no overclock potential though on my rev... For benching I use the 2Gb CSX Diablos PC16000C8 ( micron GTR) and the 2 out of 3 6gb OCZ Blades PC16000C7 (Hypers)
i'm stuk with dominator 1800 cas7:eek:Quote:
Originally Posted by Leeghoofd
until i get my other set of hyperx 2000 cas8 to run 4x1:D
i've tested a few i5 kits n :down:
i dont beleive in stock speeds:p:
I understood 4GB of ram would perform better than 2GB but how much exactly?
Only one way to find out, took a second kit of Diablo's (got 3 kits :)) and ran 32M.
I did not use any kind of tweak on both 2x1GB run and 4x1GB run, both runs were done right after a reboot.
Be prepared...
2x1GB CSX Diablo untweaked:
http://img.photobucket.com/albums/v6...Buntweaked.jpg
And here's 4x1GB CSX Diablo untweaked:
http://img.photobucket.com/albums/v6...Buntweaked.jpg
I was well surprised how tight i was able to run 4x1GB in this board:
http://img.photobucket.com/albums/v6...Msub113969.jpg
Unfortunately, i could not get the 5:8 divider to work at ~600FSB but all in all still very nice.
This proves yet again what a monster this board is!
Oh, did you guys see that k|ngp|n broke the 3D mark 01 WR again?
Guess which mobo? :D
Nice Zeus, damn weird 4 x 1gb is ~6 secs quicker than 2 x 1gb...........wonder what's causing that.
All your 32M benching is making me want to dust off my REX again, and I just got everything all nice and tidied up too heh.
Nice effort from Kingpin, the REX is still a really great board.
Thanks mate.
I reckon it's due to bank interleaving.
I had no idea the difference was this big till i tried. :)
Now dust off that REX and get to benching man, it's the best board i've ever owned!
just do the same run with eg 2 x 1 Gb and 2 x 2Gb , you will get far better scores clock for clock in PI32 with the 2GB dimms.... same for eg for S478, do pi32mb with 2 x 256 or with 2 x 512Mb, world of difference. Even Memtest gives far better bandwith with 2gb dimms over 1gb ones...
just got my other set of hypers
weird board not likin 4x1
any tips:shrug:
You could have told me earlier man. :DQuote:
Originally Posted by Leeghoofd
Forgot to check bandwidth with everest, surely 4Gigs will show better bandwidth.
The problem with 4 modules inserted is it cannot go as high as with 2 modules.
Still DDR1800 7-6-5 is pretty decent in my book but DDR1920 should have been awesome for 32M calculation.
I am back on 2 sticks again, one of the modules of the second kit needs 2.12V+ for errorfree operation at my daily 10x450 1:2 setting.
This current kit needs just 2.02V so i don't want to overflow them with unneeded volts.
Yeah my Corsair 1866C7GT kit goes quite well on this board, topped out in SPI32M testing at DDR3-2000 MHz 7-8-7-X (I didn't test higher). 600 FSB / DDR3-1920 MHz was also a good setting. Sadly I could not compare to 1GB modules as I don't own any.
@Zeus
nice OC , which is better using 4x1gb mems or 2x2gb? for Rampage extreme
ill be using this mems , but which one will i choose ?
http://www.teamgroup.com.tw/teamgrou...eem-DDR3-1.jpg
link http://www.teamgroup.com.tw/teamgrou...s.php?pl1_id=1
Thanks. ;)
I think it depends on the kind of IC that populate the modules. 2x2GB with Micron D9 chips doesn't seem to overclock well on this board from what i've read (haven't tried 2x2GB myself) but some get very nice result with 2x2GB Elpida Hyper based ram.
Personally i think good Micron D9GTR absolutely rocks on this board but finding some is like a needle in a haystack nowadays.
Don't know that Team ram you linked, so i cannot comment on that, sorry.
Zeus, any luck with the 0121 bios? I seem to be having the same problem as you and as well as boot issues. I'm thinking there might be a trick to the settings maybe?? or maybe it is a corrupt bios?
EDIT, fixed my boot issues. It was my fault, I used the wrong settings hehe.
Nope mate, i gave up on 0121. I also had bios 1104 modified but it had the same problem, no way to change the CPU multi.
I also found that the settings i used for my daily usage refused to boot with bios 0121. :shrug:
Guys,
Today is a sad day for me. :(
My golden Rampage Extreme died for no appearent reason. :shrug:
I was just browsing the internet when a little rattling noise caught my attention.
On closer inspection it seemed to come from one of my fans on the rad, turning the rpm up by the fancontroller got rid of the sound.
So i took the side panel of to investigate and then after trying to move the mouse pointer later it hung.
Still not paying much attention to it i turned the pc off by holding the powerswitch for a couple of seconds and tried to find where the ratling noise came from.
Then i turned the system back on, it did turn on but turned off by itself after maybe 10 seconds.
Oh noes, what's happening?
Tiurned it back on but it immidiately turned off again in a split second.
Now panic set in, wth??
My first hunch was that the psu must be shorted and turns off by it's short circuit protection.
Went upstairs to get an old mobo, plugged in the atx connector and jumpstarted the mobo -- psu started with everything connected to it.
Then i took everything out, all usb devices, hard disk, optical drive, soundcard, the lot.
No joy, it just starts but the onboard leds are off, only the buttons at the lower end of the board are lit, it turns off after some 10 seconds, then back on again for 10 seconds, off again and so on.
No beeps from the speaker, nothing on the lcd poster other than "welcome" .
I even took the videocard out and one stick of ram, no love...
Took the cpu out...systems starts, no leds on, that's all.
Took the E6300 Conroe cpu out of the wife's pc, same behaviour as with my E8600.
Any suggestions guys? :(
I got myself a pretty decent cpu and i want to run 600+ fsb 24/7. is this mobo capable of doing that?
Mine isn't anymore. :D But it sure as hell was!
thx for the info :)Quote:
Originally Posted by Zeus
i did prime for 5-10 min @ 600fsb no errorsQuote:
Originally Posted by kamieldehond
but i doubt the board would last long that high
so what board would you advise me then? my dfi ut p45 is a pos so i need something newQuote:
Originally Posted by zsamz_
hey Zeus, sorry to hear that bud :(Quote:
Originally Posted by Zeus
hope it's ok because your Pi and clocks are rocking on this setup.. :up::D
sounds like N.I.C.'s board earlier in the thread.. no leds then switching off.
don't know if the problem was solved, think it was a new board from shop.
anything if you use other BIOS ?
also may not help but if tweakit is still working what does LCD show for set voltages and fsb before you switch on ?
till now only the rex has gotten me priming @ 600fsb all the rest just validateQuote:
Originally Posted by kamieldehond
if your hoping to be @ 600fsb 24/7 you live in an igloo or you smokin crack
the absolute best board on 775 platform is the rex
i'm extremely surprised you were not happy with the ut p45
As cheapseats said... No beeps, no lights, except the buttons at the lower end of the board, as you said Zeus. I just sent it back to the store as an RMA, and they sent me another REX. The only different thing in my case was that it was a DOA board, from the very first moment i powered it on, it wasn't boot.
sorry to hear that mate:(, i know that feeling. If your REX is under waranty, try RMA it :up:
Cheers guys.
After reading your post, i can confirm the sympthems are identical, i can only conclude the board went belly up.
I'm pretty sad about it as you could see it was a pretty good one, been up to 645FSB at PL9 1M stable and stopped there.
I am sure there was more in the tank but decided to leave that for another day. :(
I'm gonna build the heapipe assembly back on (if i can get the small sinks of :D) and return the board.
After all, it wasn't my fault the board died, it just did.
My only concern is wether it took the cpu as well, that would be another kick in the balls as the cpu was a nice one either.
If the cpu turns out to be dead too my S775 are over i think...
I have a nice AMD Phenom II 965 C3 revision here to keep me over till i recieve a new board. Only thing i need is a board but that is already taken care of. :up:
cant get my 4x1gb d9 stable above 500 fsbQuote:
Originally Posted by zsamz_