Printable View
Obviously I didnt do that, Can you tell me why you should do the above?Quote:
Originally Posted by theonlybabyface
surely your previously saved oc profile would just load into the new bios, as it did. should I reflash and do the above? Do you really think that will make a difference?
I just find that if there is any weirdness after a bios flash, this fixes it. No need to reflash, just load up the systems defaults, save & exit and then after the reboot, go back in and load up your overclock. Even better yet, manually enter in your overclock again...
....Quote:
Originally Posted by MatroX13
theonlybabyface is right, but I suggest you take it a step further still...
Always load BIOS defaults save & reboot before you flash to a new BIOS. After, go in and either recall your saved OC settings, or manually re-enter your OC settings again...
In terms of my settings, I will list all of the ones I have changed from default configuration here...
Ai Overclock Tuner - Manual
Multiplier - 21x
BCLK - 200MHz
DRAM - 1603MHz, 7-8-7-20-1N (in BIOS, 1N == 1T)
CPU Voltage - 1.39375V
QPI/DRAM Voltage - 1.35000V
DRAM Voltage - 1.65681V
APIC 2.0 - Enabled
ROG Logo - Disabled (irrelevant but annoying :p:)
That's it. That's all I use for my OC. Everything else is set to whatever the defaults are, including Load Line Calibration (it is on Auto by default and I left it at that).
MatroX13 - Yes, there are plenty of people that have installed WC on the board, and it installs just fine. IIRC clearance for sound cards in the top PCIE 1X slot may be an issue, but I am not 100% sure on that. I am personally not cooling with water, but if you read the whole thread you will find the information you are looking for and even some examples of people setting stuff up.
CPU Clock Speed 4095 Mhz
Asus 4870x2 at stock speeds 750/900
3DMark Vantage P16819 / Cpu-25028 / Gpu-15162
Bios Version : 903
----------------------------------------------
AI overclock - manual
OC from CPU level up - auto
OC from Memory level up - auto
CPU Ratio setting - 21
CPU configuration 21
CPU Ratio Setting - auto
C1E Support - disable
Hardware prefetcher - enable
Adjacent Cache line prefetcher - enable
Intel Virtualization tech - disabled
CPU TM Function - disable
Execute Disabled bit - disabled
Intel HT Technology - enable
Active Processor Cores - all
A20M - disabled
Intel Speedstep tech - disabled
Intel Turbo Mode tech -disabled
Intel C-STATE tech - disabled
BCLK frequency - 195
PCIE frequency - 100
DRAM frequency - 1564
UCLK frequency - 3127
QPI frequency - auto
Dram timing control
1st iformation - 8-8-8-24 - 1T
2nd information - auto
3rd information - auto
EPU II phase control - full phase
Load-line calibration - enable
CPU differential amplitude - auto
Extreme OV - disabled
Current voltage
Cpu voltage - 1.45
CPU PLL voltage - 1.81592
QPI-Dram voltage - 1.375
Current voltage
IOH voltage - auto
IOH PCIE voltage - auto
ICH voltage- auto
ICH PCIE voltage - auto
Current voltage 1.674
DRAM Bus voltage - auto
DRAM REF voltage - auto
Debug mode - string
Keyboard TeakIt control -disabled
CPU spread spectrum - disabled
PCIE spectrum - disabled
CPU clock skew - auto
IOH clock skew - auto
Notes: I can run @ 4.2ghz with H.T. on but Vantage and prime 95 fails. With H.T. off Vantage and prime 95 passes. (Max 58c / Min 30c idle H.T. on)
However the Vantage score is lower by 700pts than @ 4095 mhz with H.T. enabled. :eek2:
OMG... FLASHING TO THAT BIOS NOW !Quote:
Originally Posted by dejanh
Thanks !
I cannot do 4.2 with Turbo and Ht on. My chip need 1.488v just to do 4.025ghz prime stable with HT no turbo....I need 1.544v to do 4.2 with turbo no HT.....My chip just isn't as good as the ones you guys have, I can only dream of the low volatges you guys are getting.....:down:Quote:
Originally Posted by Kensek
I use watercooling, but the NB block probably won't let you put a sound card in unless you shave some of the acetel off...... Besides watercooling the NB on X58 boards isn't really necessary in my opinion...Quote:
Originally Posted by MatroX13
Max 58c??? On all the cores?? Wow, I get about 68c at max!Quote:
Originally Posted by glenn37216
Quote:
Originally Posted by jasonelmore
These cards can take alot...
I think your Performance no ppu is throttling heavy due to v-core over the limit.., try lower both clock and v-core and result will be better.. :-)
I cant say it to many times.., these chips are heat factorys..! :)
Here is 2 air runs with GTX 295 PPU disabled...
http://i41.tinypic.com/143luv6.jpg
http://i44.tinypic.com/2pqw2nt.jpg
I will flash 0904 (on 0903 now) and give it ago.., thx for sharing everybody!
I'm wondering if more voltage on the cpu would help me get stable at 4.2. If I leave off H.T.. It can pass vantage. You guys think HT needs more voltage on the cpu ? I know it with it on it does make the cpu warmer. Whats the max voltage you guys think is safe? 1.45-1.48v? Should I go higher if the temps are resonable ?Quote:
Originally Posted by BringerOdeath
:shrug:
Quote:
Originally Posted by glenn37216
You wont c if temps are resonable.., in blink of an eye its throttling.. one easy way of checking whats goin on is running wprime 1024 and let realtemp show the massive heat...
My 920 could hardly handle 1.5 v-core cooled with SS.., over that and throttling..!!! You must cool these chips!!!
HT is the purpose of these chips.., leaving it off kinda rips initial intention to pieces...
Try finding sweetspot for your cooling and your chip.., trying to force high clocks by cutting the legs of chip.., no.., wrong path I belive...
When you find whats best for your chip.., then the nice reults will come.., and they will be stable.., takes sometime to figure it out.., but who said it should be easy to get massive clocks just like that..?
OC is not always that easy.., it takes experience and patients.., in time.., it comes.., and there is nothing so rewarding as when you actually find your way.., and can get decent stable benchmarks.., so.., dont think that your chip is no good.., give it the chance it´s worth it...
I am getting the code BCcode 124 (BSOD) when I am at 4.2 with HT turned on . Oddly this is releated to the xfi sound card. I think I'm topped out @ 4095. Temps look good at 4.2 but Vantage locks up on me. I guess I'll hold off a wile to see what others 940 users come up with. Its got to be my voltages...I just dont know which ones atm.
Im priming right now and I will keep it @ 4.2. But i have to disagree with you. If you are gaming/general Windows Vista usage the 8 threads do absolutely nothing. I think Vista 64 runs snappier with the 4 cores. Stable for me is doing @least 4 hours of Prime95 blend. Im not the 24 hour kind......If you have speciaal needs like development then 8 threads may be interesting. I can prime @ 4.2 with turbo and ht on but i think the temps are to high and the 4 extra threads give me nothing but higher temps and higher power usage..........Quote:
Originally Posted by dejanh
well, with ht on @4.1ghz no turbo, i get a max of 75c with 1.52v. I have the cpu on a single loop with a 480mm rad.
@ that voltage ...those temps are not that bad . I'm having to run 1.44-45v to stay at 4095mhz.
BringerOdeath , what is your CPU PLL voltage and your QPI-Dram voltage ? I think those voltage are keeping me from 4.2 stable.
I posted about this maybe here, maybe somewhere else before, but if any of you are using X.M.P. profiles for memory do check your QPI/DRAM setting in BIOS and then actually measure the voltage on the board. If it is not the same, then your X.M.P. is not working correctly and could cause instability of the system at which point you need to stop using it and manually configure the memory the same way that X.M.P. would do it for you.Quote:
Originally Posted by glenn37216
Manually set your qpi/dram voltage to 1.4 and cpu pll to 1.42. See if that helps any...I have my qpi at 1.45v.Quote:
Originally Posted by glenn37216
isn't CPU PLL at a default of like 1.81 or something?
yes it is, i think he means 'ioh' [nb]
Ah that would make much more sense... by the way Jason if you read this can you post your current settings for the 4.35ghz your rocking over there? I scanned the entire thread and couldn't find your settings for that... Would be much appreciated.
actually it cpu pll defaults at 1.181
No its 1.812Quote:
Originally Posted by BringerOdeath
woops, then i meant 1.842.....LOL, sorry about that
Hey guys! I have a question about OC the 965 EE. Right now I just leave the BCLK at default 133Mhz and simply pump the multiplier up, talking about the simplest and laziest OC ever. My questions are should I go for a higher BCLK/lower multiplier OC instead? Are there any real benefits, performance-wise, raising the base clock?
Tried to flash 0904.rom, loaded up the systems defaults, save & exit. Back in the BIOS, go to EZ-flash, loaded the 0904.rom from HDD this gave me the error, bad boot block :shrug: still on 0805.Quote:
Originally Posted by nordien
Also people with the cold boot issue, try to raise you're IOH voltage to 1.31216V and see if it helps, since i did this i have had no cold boot issue for 2 weeks now and counting :up:
Quote:
Originally Posted by OcAddict
Place an clear CMOS, and load the 904 from Asus ez flash 2 with usb key :)
Still stuck with this vr overide thing on this board and it REALLY sucks. Someone suggested 'lifting' the 'isense' resistor on the board to remove it permanantly, does anyone here know the location of the resistor? btw, I dont recommend anyone try it, I will try it but need to know exactly which one it is, if anyone knows please post here. thanks.
Well, if it is somehow possible to connect another resistor in series that should work as increasing the resistance increases the current limit threshold, or should anyway.Quote:
Originally Posted by bustamove44
But, I'm not up for soldering on my board ATM. :p:
put you cpu under water or something. I've never had that stuff happen to me. even with 1.55v going through my cpu. My turbo never throttles backQuote:
Originally Posted by bustamove44
Jason, lots of ppl are having the problem, basically it makes the turbo boost unusable for any kind of stabilty, there is definitely a throttling occuring.
Maybe you have a rare cool running cpu or something, mine is under water by the way and many reports are that even subzero cooling makes little difference.
I have a cascade waiting to be regassed which may or may not resolve the issue for me, but will not resolve the issue for many others who are running good air cooled set ups, why shouldnt they be able to use the turbo boost option too?
This board is the only x58 board not to have a vr overide option independent to turbo boost. all we need is for the option to be made 'indepenent' from turbo boost, as it is they are linked which makes no sense whatsoever?
Im not sure if you did already Jason but can you put up some cpuz screens of your 4.35 oc. and the temps you have. thanks
Anyone runing the 940 @ 4.2 + I would appreciate it if they would post their voltages/settings. I can get benchies at 4.2 but I get a bsod error codes relating to my sound card every now and then . I'm thinking I need a little more volts but I'm not sure exactly where. I'm a little parinoid about bumping my cpu voltage above 1.45...
My cpu will instantly throttle with 1.55v(actual 1.544) no matter how low the temps are at 4.2ghz with turbo on. 4.2 or 4.3 with HT off doesn't throttle even with 1.55....I run 4.1ghz stable with HT on no turbo(19x216 multi) fine. I think it might be the tdp limitation. What did you have the qpi voltage on when you were using 1.55v?Quote:
Originally Posted by jasonelmore
Wow , Think you may have helped me get stable @ 4.2 ...H.T. is on also. :up:Quote:
Originally Posted by BringerOdeath
-prime for 5hrs no errors ..... YET.
:clap:
Great Glenn! On the CPU/pll I meant 1.842, lol, But I am sure you figured this out....:up:Quote:
Originally Posted by glenn37216
Interesting bit of testing today. I ran wprime 32m with 1.5v [bios] and 1.45 qpi, ram at 1.69v
Tubo boost and ht on.
http://i33.photobucket.com/albums/d9...mesub6secs.jpg
The more intesting one was 1024m, did not crash.
http://i33.photobucket.com/albums/d9...e192secs-1.jpg
Load temps are high, my water set up needs attention, so that could become lower, either way, I wouldnt want load temps any higher than this, the platform runs hot for sure, but the chip seems to be able to handle it.
I'm now unsure on the throttling theory which in earlier posts was convinced was happening, vr overide still needs to be independent from turbo boost tho imo.
this was with bios 0903. the newer 0904 bios didnt yield this kind of result for me.:up:
Flashed the 0904 from USB stick and it worked :clap: strange it did not work from the HDD, everytime i flash a new bios i do it from the HDD. 0904 was not willing to do this :confused:Quote:
Originally Posted by If6was9
Now back to 0903, i find it to be better.
I finally got my i7 920 stable @ 3.8GHz HT on, turbo off but i needed to set CPU and IOH clock skew to 100 otherwise BSOD after 30 minutes in Prime95 Small FFTs, need 1.45Vcore to do this :eek:
My new 920 is on the way :D
Quote:
Originally Posted by bustamove44
Nice......... but your NB frek very low ...try higher than 3500 on NB frek..:D:D
Will do very soon, Was just testing really to get stable '1024' run I was looking for after early issues with turbo boost and ht enabled over 200x auto [21] crashing constantly.:up:
Nice Result with the Phenom II btw Hazzan.:)
Hi All
From Gigabyte I moved back to Asus and it was a wise decision! The Asus Rampage II Extreme is a real good overclocker and a high quality board!
On my Gigabyte GAX58-Extreme I was not able to keep my 940 stable @4100 with HT and Turbo off. Now I am @4200 with no issues so far, thanks to you guys and Asus!
Will do some more tests...
So Anyone know how to fix the Cold boot issue?
Im atm completely windows stable with no issues, primed for 10+ hrs, 3dmark06 looped for 3 hours, played games for awhile now, but I still get the occasional cold boot issue. I had the same exact problem with Asus maximus formula awhile ago. Dont know how it got resolved. One thing I noticed is not many people are touching the IOH and ICH voltages, I myself havent either, just because now that the memory cont is on the CPU touching the Northbridge voltage doesnt make much sense, but I have 3 GTX 280's and think that something might need a boost, any ideas?
I dont know if someone asked already, but are you using the same psu as on ur formula? also tried your mem sticks one by one to see if you get cold boot with either? or none?Quote:
Originally Posted by btdvox
Never had that issue with my rampage II. Could be to do with 3 cards maybe, asus did release a bios with a fix for ati cards that had an issue? have you checked the asus forum?
Sorry Im not, I had the Maximus awhile ago- i think about 1.5 years ago. I have a TT 1200W PSU now. Many people are having this issue, It seems to be a bug with OC'ing and the motherboard, It really has nothign to due with an unstable overclock. As I stated and you can go back one page to see my pic of my prime. My OC is very very stable.Quote:
Originally Posted by bustamove44
Im now @4300 (21 x 205) with 1.54 VCore and 1.399 VTT/QPI, Vantage and Game Stable. Not tried P95 yet.
With this MB all runs smoother as it did with Gigabyte X58 Extreme. I guess i just had no luck with that board...
One question I have. Does anybody tryed the 22 multi with the 940. If I set that multi I can't even boot above 3800...
I use Bios 904.
Thanks!
Guys, for those that don't know yet. Asus has an Official 0903 BIOS on their site. Again, that's official and not beta. I just downloaded and flashed and everything seems to be fine.
Updated Front page with Official 0903 From Asus FTP. Thanks T
NOTE: If the link takes you to a page called that says "Service Unavailable" just try again. It took me 2 trys but that may be because the server was busy or something.
Will try 903 this evening with my 940 and the 22 multi.
Added 1001 Beta Bios to front Page.
Guys check it out for me. See if it fixes man of the problems your having with VR Overide
I'll give this a go and report back. hope this is the one.
Yeah! Bios 1001 fixed the 22 multi problem for my 940. Still testing more...
Still having the cold boot issue. I'm starting to think it will never get fixed. :down:
Okay, I'm missing something...how do people get their 920 to have a multiplier of 21? The max I've been able to set on mine is 20. What am I missing?
Quote:
Originally Posted by jcniest5
Some Manufacturers boards allow for turbo to be on All the time, even with a 100% CPU Load. Asus boards, usually will not do this. Gigabytes do.
TDP Issue not fixed :(
oh no Asus...
Is that what Turbo does then set the multi to 21 ?Quote:
Originally Posted by Kensek
yes 21x on the i7 920Quote:
Originally Posted by andyr(uk)
23x on the i7 940
25x on the i7 965 when the CPU ratio is set to Auto.
Though you can set the cpu ratio up to 40 on the 965. But when it is hard coded in BIOS turbo does not engage on a 965 when set manually.
First 0903 and 0904, and now 1001. We are getting there. I hope that within a couple of weeks the TDP unlocked bios will come.
Maybe they are listening...............
I have to say that bios 904 helped me get my memory stable at 1600mhz with a 4.2ghz overclock.:up:
I'll try Bios 1001...Every new bios gets better.
I certainly hope so, this will help a great deal..the humble 920 needs it big time!Quote:
Originally Posted by nordien
bios 1001 has shown no noticable difference to me so far. more testing tomorrow.
Here are some first results from my 940.
This is with bios 1001. still fine tuning to lower vcore...
Make sure you have turbo mode and speedstep enabled in the bios. If you have disabled either of them your turbo will not work at all.Quote:
Originally Posted by jcniest5
Getting ready to order the RIIE and had a question about cooling.
Would there be any benefit in getting a waterblock for the NB? EK has this one for the RIIE: http://www.performance-pcs.com/catal...027da8257fd540 Or would it be more trouble than its worth?
Thanks in advance.
arggghhhhh
update
after wasting calling everyday phone bill around USD100 already to china/taiwan
they sent me this
ok so i called them backQuote:
Dear Kavin,
We have duplicated this issue and still no trouble found(please refer to attached file).
After confirmed with our MB designer, we didn’t put this option in our BIOS on purpose.
This function has no relation with over clocking but just a protection way to prevent power phase component damaged.
Due to our R2E’s power part is designed as 16 phase, the power part is steady enough and no need to considering the damage,
So we didn’t put this option in our BIOS because this function is no use for this board.
Here is some information for your reference:
The limits hard-wired into the core are 100 A and 130 watts.
Only the Extreme Edition allows the user to change these values at will, effectively circumventing the protection mechanism altogether.
Reference: http://www.tomshardware.com/reviews/...em,2057-7.html
http://i287.photobucket.com/albums/l...stkl1/Pic1.jpg
http://i287.photobucket.com/albums/l...stkl1/Pic2.jpg
http://i287.photobucket.com/albums/l...stkl1/Pic3.jpg
Thanks.
i told them they are stressing only 50% of the cpu 4 threads.. and around 2 cores only with single threaded orthos.
and told them to check the rule of inte turbo mode
which states something about it being enforce when stressed above a certain numbers of cores/thread only the restriction applies
and also asked them now just to indulge me just sent to me one beta bios that has the tdp/tdc limit disabled by default before i spend another 100USD
on phone calls
already been screwed over with coolitsystems for usd80 for a bracket that doesnt fit which in the end those jokers send me the fixup with canadian mail during the chinese new year which most probably take a month or 2 to reach malaysia....
will update u guys more.. wondering now should i start a donatation fund for this issue.kekekeke.
So far the 1001 bios helped me get DDR1600 stable @ 7-7-7-20-1T. This wasnt possible with the other biosses.
Attachment 93216
Just kidding??!?Quote:
After confirmed with our MB designer, we didn’t put this option in our BIOS on purpose.
This function has no relation with over clocking but just a protection way to prevent power phase component damaged.
Due to our R2E’s power part is designed as 16 phase, the power part is steady enough and no need to considering the damage,
So we didn’t put this option in our BIOS because this function is no use for this board.
we definitely need this option for high overlocking on i7 920/940.
Asus i hate you for this...
All other Motherboards in the world have this option...
cstkl1, thanks for your efforts man, Im totally amazed at that reply from Asus, I kind of see where they are coming from, but like you said, only 4 threads [1 per core] is being stressed in their test. As far as I know wprime when 8 cores are selected in the settings stresses all cores + threads in one application, hence the reason I used it in my post above.
If you do call them back, ask them if they would run wprime 1024 test with the voltages they have set above. In my opinion unless they have a magic cherry picked cpu it will not pass at those settings, are they running those tests under normal 'ambient conditions? or sitting in a fridge?
They are also running CPU-Z which does not always show the transitions properly. They need to run RealTemp and look at the clock speeds on the top. Currently, RealTemp is the most accurate tool for temps and clocks for i7.
cstkl1 - Bend the LGA775 hold-down bracket by hand and you can fit it on the Coolit FZE. I have done that right now and I'm using the block on LGA1366 with the LGA775 hold down.
I knew that was gonna happen. That was the answer I was expecting. It was probably a result of people being pushy while they were busy with other pressing matters so they just nixed the idea up front.
I basically said that in the other thread about this stuff. I said give it a few months. They are busy with compatibiluity issues and are optimizing stuff right now. I said if people didn't give it time that they'd just say "If you have a Extreme Edition CPU this isn't a problem.", and it's almost word for word.
really will give it a goQuote:
Originally Posted by dejanh
one question
the problem with my fze was it was using the older waterblock .. the thicker one..
did u try this on the thicker block or the new thinner one which the bracket officially supports??
i am not going to rest until this issue is solved and definitely before half the worlf goes off for chinese new year starting the whole of next week.
i know that email correspondence takes slowly and hence y i am calling them directly and they are replying to me via email.
dejanh and t_flight...
pm me ure emails so i can forward u all the emails that i got from them
take note guys i mean they do realize that there is some issue and really wants to make this board the best board.. they are onboard on this
but theres a communication issue with them cause of their grasp of english.
dont mind asking my gf to talk to them in mandarin but that would make things worst as the translation would be lost cause she wont have a clue what i am talking about.
so if u both of u can email him
and with me calling him. hopefully they will just give in and send one bios that has the function disable on default like what gaygaybyte did.
bro i seriously think and somehow know from talking to themQuote:
Originally Posted by T_Flight
first they didnt realize there was such a issue
and they are not replicating the issue correctly
and they dont think this issue is hindering overclocking gains
appreciate it if both of u can email me ure adds so i can forward all the mails
and hopefully with much clearer communications from u guys and with me on the phone.. something will get through.
It's nice to see a line of communication has been established. Can you PM me the email and phone number you've been calling which has a direct line to the R&D department? Also if you don't mind, forward me those emails also.
In the meantime I'll keep you guys informed of the latest bios's, BETA and Official as soon as I get my hands on them.
jasonelmore@yahoo.com
jasonQuote:
Originally Posted by jasonelmore
already emailed u everything i had that was relevant in one email
take note this was using mozilla thunderbird which i just did forward attachment
some of the dates and a lot of repetitive but the last email if u scroll down was the one i posted above.
also this is the global tech and engineers.. the r&d .. no direct contact yet.. but i guess if i really want can get them but i am worried about the communication issue which u will see after u speak to them
their contacts and emails in on their sig but i advice emailing them first and imediately calling them afterwords..
reasoning is just to provide them in words the issue u want to address and also a form of contact to reply to u.
calling them just to double check and heck i can always get this guys.
call them during +8 GMT 9.30 a.m. to 5p.m.
on another issue...irrelevant but amusing
coolitsystems..
their staff seems to be working only from 12-1 p.m. to 5p.m.
business must be good to be so relaxed.
Alright thanks. It got flagged as spam for some reason. But i've got it. Thanks again
Great job getting in contact with ASUS man :up:Quote:
Originally Posted by cstkl1
Are they aware of the cold boot issue so many people are having including me? Just wondering ;)
I'm on the 1001 bios right now but i don't see any big differences, still can't boot @ 200 x 20 with my 920. 215 x 19 on the other hand will work but isn't stable. Like is said before i've got a bad 920, new one on the way :D
Guys, can I overclock my i920 if I just have PC3-1333 memory? It seems like you guys all have PC3-1600 for those extreme overclocking...unfortunately I couldn't afford it so I settled for only PC3-1333. And if I can, please advise the BIOS settings. Thanks...
cstkl1, I sent you a PM...let me know if you got it. I also included some details on what I did to the FZE to make the LGA775 bracket fit LGA1366. I will post some screenshots of my setup later too.
Flashed 1001 , I still cant POST at anything above 196 bckl with any multiplier , voltage, uckl and ram dividers . Running out of options here other than wait for bioses and settle for 4100 HT on which is prime stable for me at 1.39vcc ; 1.35qpi/pll.
Quote:
Originally Posted by ACS
Volts are way too low.
Qpi Pll can go up to 1.65V
and CPU can go to 1.55V
Thats my 4.1 stable clocks , ive tried cpu up to 1.65 and qpi to even 1.69 just to check if it will post , dram up to 1.8 aswell. No post at all, lowest qpi and uckl dividers + laxed mem timings
doneQuote:
Originally Posted by dejanh
as again.. really want them to blindly sent us a 0903 beta modified bios with the tdp/tdc disable by default if things are again not clear.
before the cny holidays in taiwan..
btw did a big complain to coolit
but really would welcome a pic..
thanks ..
haha coolit is on your radar man. spam em until you get things fixed
I don't understand how they don't see a connection between a 100a/130w limit and not being able to oc? Do they not realize that these cpus eat power like there's no tomorrow? My chip runs into issues even before 1.3v, because of the heat and the subsequent high tdp. I don't care if they only give us a beta bios which won't be covered by warranty, I just want to oc my chip in peace. If not, then I will get the new evga board, high build quality without the bios policing...
Hi guys!
Thanks for all the good info fellas. I'm stuck at 190 blck 20 multi with turbo on, HT off (LinX stable). Any ideas on how to to reach 200 blck with turbo on?
Here's what I'm running...
920 (batch 3836A984)
RE2 bios 805 (tried 904 but didn't see inprovement)
Corsair HX1000
Corsair Dominator 3x2gb C8
Custom water cooling setup 2 loops
http://www.aquariummasters.com/u.jpg
http://www.aquariummasters.com/1.JPG
http://www.aquariummasters.com/2.JPG
http://www.aquariummasters.com/3.JPG
http://www.aquariummasters.com/4.JPG
http://www.aquariummasters.com/5.JPG
I've tried everything I could think of....vcore up to 1.52, vtt to 1.45, vdimm to 1.70, ioh = 1.40, ich = 1.20 but nothing.
I can post at 200 bclk but it BSOD on Vista splash screen. Highest I could do is 197 x 21 into vista but 5 secs into LinX it crashes.
Any tips or info will be greatly appreciated.
Thanks in advance:D
Lou.
Hey,Quote:
Originally Posted by Supafish111
Ok, I can tell what you are messing up right away so I'll have you try something...
Set your settings like this and then try...
BCLK 200MHz
Multiplier Auto
DRAM 1600MHz, 8-8-8-24-2T (to start, tighten later)
CPU Voltage 1.42500V (to start, increase if needed)
QPI/DRAM Voltage 1.35000V
DRAM Voltage 1.65681V
Remove the 700mV differential amplitude and set all other settings on default including LLC.
Try to boot into Vista and see what happens...
Thanks for the quick response dejanh!
Followed your settings exactly but was unable to get vista bars to show then it reboots. Tried vcore up to 1.46 in bios....show I go higher?
Also, QPI voltage shows 1.323 even when its set to 1.35. Does yours do the same? On timmings page, am I right to assume 2T is the first item on 2nd info (2N)?
Thanks again!!!
Lou.
Hey guys,
Anyone running dominators C8 care to share their timings? Thinking memory might be the culprit holding my 920 back.
Thanks!
Lou.
No, the QPI/DRAM should be reading pretty much exactly 1.35V when it is set at 1.35000V. Mine reads (measured with a DMM) 1.349V.
I am running the Dominators and I have them set at 1603MHz, 7-8-7-20-1T, 1.35000V QPI/DRAM, 1.65681V DRAM.
Currently running my low voltage OC on my i7 940 at 3.8GHz (19x200MHz) and 1.20625V on the CPU in BIOS.
Have been stable at 4.2GHz with 21x200MHz using as low as 1.39375V in BIOS. Normally I would run 4.2GHz using 1.42500V in BIOS.
Could be the chip, or could be the board...not sure...
Can you take a DMM and measure your set voltages on the board to make sure that they are approximately what they are set at and not undervolting significantly? Maybe it is your board that is causing problems...
Supafish111, have you tried Bios 1001, worth a shot?
I will tell you that i had a tough time getting my memory stable at 16000mhz with a 4.2ghz overclock, then i flashed 904 and boom!!! stable as a rock
Here are the readings via DMM
BIOS Setting DMM BIOS Reading
vcore 1.4812 1.484 1.468~475
PLL Auto 1.845 1.812
QPI 1.350 1.344 1.330
IOH Auto 1.15 1.14
ICH Auto 1.12 1.11
vdimm 1.656 1.66 1.64~1.54
This is with LLC set to Auto. Gonna load up bios 1001 to see if it helps.
Thanks guys :)
Lou.
Anyone have alternate link to either bios 904 or 1001? Kensek's link doesn't work.
Lou.
sigh....bios 1001 didn't do much.
Shot 1.53V to the chip (briefly) and I did manage to see vista bars complete, but after that it blue screens on me and reboots. Even tried running mem on 6x but still no go. These suckers sure are hard to OC to 4ghz.
I envy all the fellas here doing 4ghz+ under 1.45vc. Gotta keep trying tho so keep the ideas coming!
Thx,Lou.
It's just than there has been 100 downloads today and that's all the free account allows.Quote:
Originally Posted by Supafish111
Try again tomorrow morning.
Has anyone experienced any sleep wake up issue with this board? I can only wake the computer from sleep with the keyboard, which is a Saitek Eclipse II, but once the computer is up and running, the keyboard would then stop responding. Device manager shows the error "Windows cannot load the device driver for this hardware because a previous instance of the device driver is still in memory. (Code 38)". I have to restart Windows to make the keyboard working again. I'm on BIOS 0903 right now.
I've upped both to Rapidshare (Premium) to ease server ;)Quote:
Originally Posted by Kensek
Code:http://rapidshare.com/files/186915679/RampageII-ASUS-Extreme-0904.rom
http://rapidshare.com/files/186915522/RampageII-ASUS-Extreme-1001.rom
thank you
Nice mobo, guys :up:
Added Alternate download links to first page for both beta bios versions 0904 and 1001
Anyone?Quote:
Originally Posted by jcniest5
Yes, you can. Just use a lower memory divider if you need to. However, I am pretty sure that your 1333MHz memory will do 1600MHz fine, though you may need to relax the timings to 9-9-9-24-2T or something and use 1.35000V QPI/DRAM with 1.65681V DRAM.Quote:
Originally Posted by jcniest5
Now, on a different note, I did some stuff on the TDP override front...
I sent a full set of detailed instructions on how to reproduce the TDP problem to the Asus people. It was nicely formatted, in numbered steps and bolded important parts. I also provided them with links to direct downloads for the tools used to stress the CPU and to RealTemp for monitoring the temperatures, clocks, and multipliers.
They replied that they forwarded the information to their engineers and will get them to test it today or tomorrow, and then get back to me.
Let's see what happens.
By the way, below are the instructions I specified...
1. Use an i7 920 retail processor.
2. In the BIOS overclock the CPU using the following settings:
CPU Multiplier Auto
BCLK Frequency 200MHz
CPU Voltage 1.48125V
DRAM Bus Voltage 1.65681V
QPI/DRAM Core Voltage 1.35000V
Do NOT change anything else in the BIOS. Leave Load Line Calibration on Auto, and leave everything else as is.
3. After settings those settings save the BIOS settings, then start Windows.
4. Download Mersenne Prime (Prime95) from this link http://www.mersenne.org/ftp_root/gimps/p95v258.zip
5. Download RealTemp 2.90 from http://www.fileden.com/files/2008/3/...alTempBeta.zip (this is sourced from the first post in this thread http://www.xtremesystems.org/forums/...d.php?t=179044) and is the ONLY program that correctly reads the Core i7 CPU temperature and speed.
6. Unzip both programs to directories.
7. Start RealTemp and look at the top. You will see the processor information there, the current clock speed, and load.
8. Start Prime95 and select "Just Stress Testing" if asked. Next select Options > Torture Test > Small FFTs and then make sure that the box that says "Number of torture test threads to run" says 8. If it does NOT have 8 please write in 8. Click OK and let the test run.
9. RealTemp should now be showing 4200MHz speed, 200 x 21. If it is NOT showing 4200MHz and 200x21 when the CPU is under load you are throttling due to going over the TDP limit. It may take a few minutes for it to show up. Also, the temperatures have to go over 80C.
If they cannot follow this then I do not know what they can follow :p:
If you lowered the RAM down to take it out of the equation and ran that high of a vCore to it and it still wouldn;t boot, it's the chip. It's gotta be.Quote:
Originally Posted by Supafish111
If you haven't, lower RAM down, and relax timings, then see what the CPU will do. That isolates the RAM out so it won't affect the CPU OC. After that, if it will OC, start raising the RAM into it, then tighten the timings to find the sweet spot. You may have to back off BClk a little too since when you bring the RAM back in it might not run at the max clocks it did with the RAM totally out of the Equation, since the BClk affects both the CPU and RAM clocks.
Check PSU voltages also with a Multimeter and make sure they are stable and don;t droop under load or fluctuate. That can raise havok with OC's. If you can also check voltages at the board. I didn't see what board you have, but on the Rampage II Extreme you can do this. Others, you have to know where to put the probes.
It is luck of the draw with some of these. Not every chip will OC the same...even in the same batch numbers. It depends on where the chip came from on the wafer, how it was binned, the date it was manufactured, and lots of other stuff.
I would check those things above though.