1083 it's enough for me in daily.
You are tempting me to get another kit and go 8gb. I want to test my modules first :)
You should really use a ramdrive with 8gb btw :)
Printable View
Can someone list their settings from the overclocking tab in the BIOS for this board?
I'm trying to get 333 * 12 to work with my QX9650, but apparently, I need to tweak the voltages and I'm not quite sure what to tweak :shrug:
After trying to fit my Tuniq Tower today i soon discovered that the Tuniqs Mounting plate is incompatible with the Rampage. It has two round bits sticking out on the end for K8 mounting which i had to saw off so that it fits nicely now :)
max fsb for now with the NB at 50C...
http://img353.imageshack.us/img353/4181/dsc02509gz4.jpg
What? You got 11.265s at 5.6GHz? That's pretty slow.
X7 multi...just searching for fsb limits M8..
It seems that I had to set these values to get the board to overclock:
CPU GTL Reference : 0.63x
North Bridge GTL Reference : 0.67x
Ai Clock Twister : Strong
Right now, I'm at 400*9 with 1.325V (1.31V under load) @ 50C load with Thermalright Ultra 120 eXtreme with 2 Scythe S-Flex fans (push/pull). My RAM is at 1066 MHz with 5-5-5-15 and 2.1V. Will push it further this evening.
Really? Mines at 400*9 with my QX9650 at 1.25v @ 33c with the Thermalright IFX-14 with 1 Noctua NF-S12-1200.
It was stable with al voltages at AUTO, according to CPU-Z it was 1,275v on AUTO so I lowerd it to 1.25.
Argh want the 4Ghz, but it's so not working at 400*10 or 444*9!
Hey guys, I've got a Rampage Formula on the way for my new rig (build log in sig) and had a few questions about cooling it.
HR-09's should fit without modification provided I get the correct ones, right? Also, I'm going to be using OCZ ReaperX, and was wondering if I used an HR-05 IFX (SLI or normal) on the NB, would there be any clearance issues with the RAM?
Thanks guys, I'm sure I'll be back here once I get this up and running for some OC tips.
It all depends on the CPU Cooler, see my post in this thread here: http://www.xtremesystems.org/forums/...&postcount=271
And of course don't make the mistakes I mad :D
CPU will be under water. I was wondering how close the HR-05 comes to the RAM slots as the Reapers have sinks which extend slightly beyond the edges of the slots.
I have read that the HR-09 near the I/O plate needs to be dremeled so it doesn't hit caps. Did you have any trouble with this?
I think my OCZ Flex's heatsinks are as high as your Reapers, so no you won't have any problem there.
Uhm am Dutch and looked the word 'dremeled' up but couldn't find it, what does it mean :shrug:
The HR-09 will fit perfectly, unles you buy the wrong size like I did. Even then I had no problem with the I/O plate what so ever.
Anyone here got a C1 revision of the QX9650?
http://img103.imageshack.us/img103/545/cpuzjr6.jpg
Still not getting it at 4Ghz stable, hell even boot.
Also I've set al voltages to the lowest, except the memory at
2.2v and I'm unable to set the vcore. CPU-Z shows 1.26, but
setting it manually to 1.25 won't let me boot.
I've been doing some research in the T1/T2 or CR1/CR2. T1/CR1
is faster but only at lower clockspeeds, it will give higher
readingspeeds. Is that correct?
Cause my OCZ Flex 1150mhz 5-5-5-18, will only run stable at
1066Mhz 5-5-5-18 2T/CR2 at 2.2v.
If you google it you might be able to find it. If not, don't cry or you'll need kleenex. I could always zerox a document that explains it.:DQuote:
Uhm am Dutch and looked the word 'dremeled' up but couldn't find it, what does it mean
just a bit away from 600fsb...the RF performs very well but i think i have a cpu fsb wall here...
http://www.imageshack.gr/files/fgb9j6agw5f4d82nyf6s.jpg
[QUOTE=giorgos th.;2981603]just a bit away from 600fsb...the RF performs very well but i think i have a cpu fsb wall here...
Can you share your bios settings for this?
Vmch @ 1.81
Vpll @ 1.58
Vfsb @ 1.54
Vsb @ 1.175
Vsb 1.5 @ 1.60
PL10,strap 333,mem 1:1.
After a week in the hospital and 2 weeks recovery I feel like I'm almost alive again. But enough about me.
I finally got my new build finished with a q9450 in a Rampage Formula. I'm concerned about my NB and SB temps.
Before I installed the mobo I removed the entire NB/SB/heatsink assembly, removed the oem paste or whatever it was, applied ac5, and reassembled the mobo. I think that I did a decent job of it. It is possible that I used a little too much paste.
Note that my ambient room temp is a little high: 30c
But I'm concerned that my NB and SB temps are very high and this is at idle.
Somewhere I read about raising the NB and/or SB voltage and how that made temps go down but I can't find that advice after numerous searches. Maybe I dont know how to search, what terms to use..
Can anyone offer some advice on this? I haven't done any OCing yet as I'm just trying to get this computer up on my lan (it is a disk farm replacing 2 old computers that are failing so I'm more interested in just getting it working than speeding it up just now.)
Here, from PC Probe II:
ambient is 30cCode:Vcore: 1.20v
+3.3: 3.18v
+5: 4.92v
+12: 11.98v
SB: 1.07v
SB2: 1.52v
FSBT: 1.17v
DRAM: 1.94v
DRAM Ref: 1.34v
NB: 1.34v
CPU: 33c
MB: 44c
NB: 53c <<<<<
SB: 51c <<<<<
Edro
oh yes....it`s about 45-50C...that`s why i have PL10...guys with watercooled NB are working with PL9 (cas5) - PL8 (cas4) there...
still on 0219
and also the 1.65Vmch is not needed for 450fsb and 333 strap....too high..
Yep - FSB vtt is way too high. Intel states absolute max is 1.45v (so does anandtech).
you can go up to 1.55 with sub-zero cooling..
I think it's the chip. I've tried just about everything....
Why is my core at 1603 and not 2.53GHZ :confused:
C1E -- disable it.
Hmm try disabling TM.
That is an indication that C1E or EIST or something like that is enabled. Pull your CMOS battery and clear the Cmos, then try disabling them again.
I have recently purchased this mobo i am trying to decide wich type of ram to get for it.
Should i just get pc6400 800MHz or pc8500 1066Mhz?
I recommend looking for some GSkill DDR2-800 HZ. That is what I am using and it OCs like mad so it will do both DDR2-800 with at least 4-4-4-12 timings and it will do DDR2-1100+ with 5-5-5-15 timings on this board. NewEgg still sells it but I am not sure if it is still D9GMH which is the IC of choice. But you can find it on fleabay and on different sites for sale and you can ask if the ICs are D9GMH. There is a GSkill link here on the forum where you can ask if the latest modules are still D9GMH.
Technically any D9GMH module will work very well covering a broad range of OC settings (DDR2-667 to DDR2-1200+). I just have a warm fuzzy for the GSkills.
0403 official BIOS is out:
http://support.asus.com/download/dow...%20Formula
Just another memory tweak it looks like. Anybody try it yet?
Anyone seen this about loadline? http://www.hardwarecanucks.com/forum...review-14.html
I thought it was a no no for 45n cpu's.
Well it's said it is a no no.I think that site is where it came from.I have mine turned off.I don't have much of a vdroop.It's only 0.02volts.Now I may find out different when I go to do some heavy overclocking.But it's still going to be turned off till I hear different.I still have both spread spectrums turned on and still clock to 450fsb.
:)
With loadline enabled on my Q9450, it is prime stable at 3.6ghz with 1.35625v, and never crashes during intensive load. However, it randomly BSODs and resets at idle.
With it disabled, I need to increase the vcore to 1.40625 to get prime stable, and it no longer crashes at idle.
I really don't see what it would hurt, just set vcore at the value you really need in bios.
I don't understand the issue either, and I sent a ticket to ASUS support about it, detailing all the info I knew. I hope it is fixed in a BIOS release.
Hey,
Tiny question: which Intel Chipset Installation Utility version are you using with the Rampage?
I installed the latest one I found on Intel's website (which is 8.3.1.1009, thus newer than those on Asus' Install CD and website) but it's not said anywhere that it actually really supports X48 :confused:
If you go through Intel's X48 chipset information page to download software, etc it will take you to the latest version of the chipset software. Based upon this I would say the software does support the chipset properly.
http://www.intel.com/support/chipsets/sb/CS-028823.htm
8.3.1.1009 9/24/2007
Not so new. The one from msi (linked before) is more recent maybe?
I use the v8.3.0.1013 Intel(R) Chipset Software Installation Utility (which appears sorely our of date) available on Asus global website here on my Vista Ultimate 64bit installation. A have also upgraded MB Bios to v403. No glarring issues but detailed system testing not yet completed.
See here for results when searching for x48 (series 4 drivers) on the Intel website. :shrug:
The links provided by 003 seem to have much newer versions, but I am unable to locate them on Intel site. 003 have you used these newer drivers you referenced? .. are they working okay for you?
Yes, I use them, as do most people. For whatever reason, intel has terribly out of date drivers on their site, they have not been updated for like forever.
Thanks for the info 003.
Has anyone experienced the atikmdag error. It seems there are a huge number of forums about the problem and that there is currently no 100% fix. I just returned my entire system for repair because it was freezing so frequently under Vista64. It was after that I discovered the numerous forums.
I realize this may be straying somewhat off topic so feel free to delete question if so.
Anyone knows what's the latest revision of Ramapage Formula? What I heard about is 1.00G only. Just confused that Anandtech in its review reports about board revision 1.03G while it is perfectly seen on the pictures that they have 1.00G.
http://www.anandtech.com/mb/showdoc.aspx?i=3208&p=2
Can anyone confirm that 1.00G is the latest revision of RF?
It's probably a typo. No way they move through the alphabet on 1.00, then again all the way up to 1.03, and then all the way to G. Mine is 1.00G.
I'm getting almost exactly the same behaviour trying to get my E3110 stable at 4GHz+, Load Line on I get no problems at full load and much less vDroop but at low usage/idle is will just freeze/crash. Without there is quite severe vDroop but no crashing (so far).
Any word on this issue?
They said it's overshooting some voltages or whatever. If all it does is prevent the voltage from dropping under load, then I don't see what they are talking about, unless ASUS has implemented it poorly.
This is the article at anandtech -
http://www.anandtech.com/cpuchipsets...spx?i=3184&p=5
Basically what they are saying is that vdroop is not a problem - its deliberate, and that the vcore you set in bios is a MAXIMUM allowable vcore, not the TARGET vcore. If you enable loadline calibration then you risk overvolting your CPU without even realising it. Read the article and it all makes sens ;)
I presume its any implimentation of vdroop control not just asus loadline calibration that it refers to. However, i used it on my DFI P35 with no problem.
Brama:
What revision of the board are you using? What BIOS? What voltage settings? GTL settings?
Not exactly opposite settings, you're using C1E where most overclocking I'd say disable it. So at idle it'll be under clocking/volting.
vDroop of .05 and greater isn't a problem? :ROTF: with LLC on at least it was a predictable and stabe .02 vDroop. Pencil mod helps a little but come on.
My P5E had no such problems with LLC though, only my Rampage Formula.
The article really doesn't make any sense when overclocking with these boards, without LLC you have to set a FAR FAR higher voltage to get stability. So much so that in my completely uneducated opinion you'd never exceed the 'max voltage' you need to set without LLC on when using LLC and the 'over volt' state occurs.
So your P5E works with LLC enabled and using a 45nm CPU? No random crashes at idle?
Exactly, both E8200 and E3110 at 3.8 and 4GHz LLC always on and 0 problems.
Both are pencil modded, did the Rampage last night and it made MUCH more of a difference than to the P5E.
Wait, so the pencil modded rampage makes LLC stable with 45nm?
Or it eliminates vdroop with LLC disabled?
What exactly is the pencil mod and how do you do it?
It decreases vDroop
As per the Maximus Formula thread. It's the same on all Maximus / Rampage based boards.
http://img128.imageshack.us/img128/4...roophs4wg8.jpg
Use a HB / 3B / 4B pencil, 2 or 3 strokes only.
What are you supposed to do? A soldering pencil? Writing? And what do you do with it? Once done, do you leave LLC disabled?
Electrically, what is the pencil mod actually doing?
OK regardless of pencil mod.
P5E - LLC works as expected
Rampage - random crashes with LLC
The pencil mod is quite literally getting a 'grey lead' pencil and drawing a few strokes where marked in the image, it's tiny. My understanding is it reduces the amount of vDroop by reducing the resistance. It did next to nothing for my P5E with or without LLC, it greatly reduced the vDroop on myu Rampage Formula without LLC. I have not tested with LLC back on yet.
Ok I see what it's doing. Wouldn't a better and more electrically sound way of doing this be soldering a resistor in parallel with that SMD capacitor?
If you can be bothered soldering something that small when you can just take 5 seconds with a pencil for the same result, go for it. Also the pencil will rub off for RMA :rolleyes:
Yeah but I heard that the pencil wears off after a while.
EDIT:
When you refer to P5E, do you mean the P5E3 or the P5E Deluxe?
Would it not be possible to flash the Rampage Formula with the P5E Deluxe BIOS?
How is the pencil supposed to wear off exactly?
When I say P5E I mean the bottom of the pile plain old no frils P5E that shouldn't to ANYTHING better than the Rampage Formula.
I heard that after a while the graphite can wear off. Not sure if it's true or not, or how long something like that would take.
The issue with LLC really bugs me. Perhaps we should try and round up a bunch of people with a Rampage Formula and a 45nm CPU that all have this issue, and submit the problem to Asus? It's more likely they will do something about it if many people complain, as opposed to one or two. I have sent a support ticket to them about the issue and as of yet, not gotten any response.
It's a strange and irritating issue especially because of the significant vDroop on all the boards based on this PCB.
I just want like 3.2 GHz from my Q9450, how hard is that going to be?
I've just tested my Rampage a bit more and with the pencil mod I only get .008 vDroop :)
I have my QX9650 @ 3.66 GHz with 1.3875V in BIOS and at load, it reads 1.37V ~ 1.36V with LLC enabled. Using BIOS 401.
That's the strange thing... I can't. Although, the previous owner was able to run it at 4.0 GHz with 1.25V, granted, he was using vapo. When I received it, it actually required more voltage to run at 3.66 GHz, so I RMA'd the CPU and got a new one last week. This one tends to overclock better with less voltage and runs a bit cooler. By the way, how in the world did you get a C1 stepping?! It's not even on Intel's Processor Specs Sheet. The only C1 stepping they have listed on their site is the QX9770(or is it 5? The LGA775 one lol)
Is everything under Advanced disabled?
It is a C1, a friend of mine is an overclocker who can get samples. I bought it from him.
But can you help me figure out what it is what I have. CPU-Z reads it as an C1 but it does not show it as an Energy Sample (besides how can you see that it is)
Here a shot:
http://img103.imageshack.us/img103/545/cpuzjr6.jpg
I'll have to check when I get back home. As far as the CPU-Z screen shot. That is very odd. What was on the Intel IHS? And I thought ES stood for Engineering Sample. Also, I'm now reading that 1.40V or more is bad for the 45nm Quads and will degrade their life.
I agree entirely with what you are saying about the VDROOP on the Rampage. I initially thought I had a crapola processor (Q6600) it could not run for 1h P95 (Blend) on Stock 2.4GHz with 1.2000v, The voltage fluctuated wildly from 1.14 to 1.152 to 0.04v. Now with the Volt Mod in place Bios = 1.200v and CPUz Load=1.184v and it does not fluctuate.
Thanks for the Pics.