UnOfficial Asus Rampage II Extreme Thread

[Dark-Master]

@ RoEy

Maybe you just picked up a not-well-for-high-overclocking D0 chip.

I've tried so hard for mine too. I have to use it @ 1.4V with LLC Enabled for 200x21 or even 200x20, but only 1.22xxxV for 200x19.

[Nuckin_Futs]

Ok, I've dropped ram down to whatever the lowest speed is for that particular bclk and also the uclk so it's double my dram frequency.

I'm now going to raise the bclk and see how far I can push the stock voltage, 1.208, currently sat at 3.9.

I'll test it using small ftts in prime and maybe IBT normal ram use perhaps? Hopefully I'll manage to get to 4.2 then when the cpu is stable I'll push ram upto 1600 and test it all.

I'm not taking ram over 1600 so I always thought I wasn't overclocking before.

Does this sound ok?

RoEy

You're on the right track so far. As far as the mobo and ram is concerned, we are overclocking even to use DDR1600 since the Jdec standard is like 1066 or 1333 for most modules. So keep in mind as we up the BCLK we need to up a little at a time to other key items like NB/SB, QPI, VTT, and at times even PCI-E freq. I do believe we need to focus on QPI when also trying to run DDR1600. Hopefully others can help add more fact to this part and how much at a time. I am using 1.35 to 1.40 QPI qith DDR1650 1800. Below is a basic settings for a 920 DO @ 200*20 DDR1600. [as set iin BIOS] = active

Voltages
Vcore [1.25625] 1.244 (some may need closer to 1.3v) I have 1 that needs 1.28750 in BIOS
PLL [1.81592] 1.819
QPI [1.26875] 1.230
IOH core [1.17966] 1.177
IOH PCI-E 1.5v [1.52431] 1.521
ICH core [1.17966] 1.177
ICH 1.5v [1.52431] 1.521
DRAM Bus [1.65681] 1.660
DRAM Ref [Auto]

Here's some basic CPU settings;

CPU Configuration
CPU Turbo Power Limit [Disabled]
CPU Ratio [20]
under CPU secion, I disabled all but Intel HTT, and then
Intel HTT [Enabled]
Active Processor Cores [All]
Loadline Cal [Enabled]
BCLK Frequency [200]
PCI-E Frequency [110]
DRAM Frequency [1606MHz] 801.6 MHz x2
UCLK Frequency [3200]
QPI Frequency [Auto]

[Nuckin_Futs]

Can I get some help from one of the regular overclockers here? I want to learn a more in depth usage and control of these settings listed below for my current 200*20 4.0GHz profile but with desire to move to a stable 4.2GHs with parts in Sig.

QPI Frequency = [Auto]
DRAM front to back timing at 1600 ~ 1800MHz [Auto] sets to 0
CPU Differential Amplitude = [Auto]
CPU Voltage Control = [Absolute (VID)]
DRAM Ref [Auto]
CPU Clock Skew = [Auto]
IOH Clock Skew = [Auto]

As for the UCLK Frequency, I set at exactly twince my DRAM ie; 3206, but is there any need or benifit to rais above and if so how much at a time o test?

Also, for the CPU PLL, what kind of tweaking are you regulars upping this above defult? I am sitting at 1.82917 for now, 1 step obove defult.

Aything else you feel I left out, please add.

[RoEy]

Thanks Nuckin for your help. I'll do more testing this week and hopefully I'll get to 4.2. I spent ages trying to get 20x200 but the 20 multi was just too unstable. I'd also like to know the answers to some of your questions.

Cheers,

RoEy

[DashTrash]

Hey Cryptik,

I saw back in post 3594 you had your ram @ over 1000mhz with about 1.3v qpi. What kind of ram are you using? I have a D0 920, R2E (bios 17xx), Gskill Perfect Storm 2000 (8,8,8 version, don't have exact model # in front of me) and have so far been unable to get to 1000. I have gotten to 952 relatively easily, but 1000 remains elusive. After some research, it seemed like it was safe to use 1.65 vqpi without killing anything, so I tried that high, along with several lower voltages. I also tried upping the Vioh, but no luck. No post. I'm at work right now, and won't be in front of my computer for several more hours, so can't post my setting (can't remember off the top of my head). Is hitting 1000mhz with ddr3 2000 kind of a hit and miss kind of thing? Is that usually a limitation of the chipset, ram, or mem controller? Is it a true statement to say Qpi voltage allows higher mem speed, NB volts (can't remember what its called in the bios) allows higher bclk? Thnx!

[Nuckin_Futs]

I am still stuck on the system not waking up after putting into sleep mode. The debug screen shows time and monitor is black screen with no signal when I tries to wake her up. I tried every BIOS I can get on and still no luck. Anyone got an idea how to fix this? Also, how do I check what revision my board is on?

Thanks in advance.

I am experiencing this on mine too since several BIOS since new. I could use full ACPI S3/S4 modes when RAM is up to and slightly above 1600~1800 but with BCLK closer to stock. But as I climb higher and above BCLK 180 keeping RAM under 1600 for testing, it doesn't stay asleep anymore.

Now I changed CPU's to a i7 930 & it enters and remains in suspend modes but locks on resume. I even wen to latest ASUS site BIOS & it crashes less but not reliable enough yet. S4 will lock requiring a reset to resume last session but S3 crashes on resume almost always now.

Let's keep eachother posted on this and see if we can't figure it out since we might be on our own here. I'm thinking it may be a volt tweak needed to cerain area but hopefully one of the more experienced will shine some light on this for us.

[CryptiK]

Hey Cryptik,

I saw back in post 3594 you had your ram @ over 1000mhz with about 1.3v qpi. What kind of ram are you using? I have a D0 920, R2E (bios 17xx), Gskill Perfect Storm 2000 (8,8,8 version, don't have exact model # in front of me) and have so far been unable to get to 1000. I have gotten to 952 relatively easily, but 1000 remains elusive. After some research, it seemed like it was safe to use 1.65 vqpi without killing anything, so I tried that high, along with several lower voltages. I also tried upping the Vioh, but no luck. No post. I'm at work right now, and won't be in front of my computer for several more hours, so can't post my setting (can't remember off the top of my head). Is hitting 1000mhz with ddr3 2000 kind of a hit and miss kind of thing? Is that usually a limitation of the chipset, ram, or mem controller? Is it a true statement to say Qpi voltage allows higher mem speed, NB volts (can't remember what its called in the bios) allows higher bclk? Thnx!

I'm using a Corsair Dominator 1866C7GT kit. I think it does seem to be somewhat hit and miss with this board, seems to be easier on some other boards. I had B2B delay still set to 4 as well, didn't seem to make any difference by loosening it off. Higher Vtt (QPI) does help both ram and BCLK, also uncore. vIOH hasn't really helped me much except for stabilizing 218 BCLK - needed around 1.13v IIRC so just a small boost. Both Vtt and vIOH would come into play when pushing higher.

[CryptiK]

Just did a little Linx run now I have my cooling sorted.

4.4 GHz @ 1.256v

[zsamz_]

nice cpu cryptik ? batch

[Mydog]

nice cpu cryptik ? batch

And Bios vcore.

[Dark-Master]

And Bios vcore.

Some of bios settings would be great too.

[xpower]

Hi extreme people
sorry guys for some nobe questions.....
i had a new system running for a month now.R2E Bios 1639 with 920 Do chip on stock HSF,volts etc.first of all with the ambient temp of 32 the cores are 40 38 41 37 on idle with core temp 0.99.5 reading.why is so much difference in between cores,first week and a half the difference was not so much,but now after one and a half month of using the differene is more noticable. is it normal? i notice the Mother Board is little bend on top where the top HS for mosfets.i used lots of asus MB,last one had same problem it bend where the cpu is and on top like this one. did any one had bend problem.is it because all MB are made in china sh**t.this board is in Antec P160 case.
sorry for my bad english.

[CryptiK]

nice cpu cryptik ? batch

Thanks it's a 3849B018

And Bios vcore.

I don't use LLC. Bios set vcore is 1.37500v

Some of bios settings would be great too.

BIOS settings:

vcore = 1.37500
vPLL = 1.81592
Vtt = 1.32500
vdimm = 1.56406v

Skews = Auto

Hi extreme people
sorry guys for some nobe questions.....
i had a new system running for a month now.R2E Bios 1639 with 920 Do chip on stock HSF,volts etc.first of all with the ambient temp of 32 the cores are 40 38 41 37 on idle with core temp 0.99.5 reading.why is so much difference in between cores,first week and a half the difference was not so much,but now after one and a half month of using the differene is more noticable. is it normal? i notice the Mother Board is little bend on top where the top HS for mosfets.i used lots of asus MB,last one had same problem it bend where the cpu is and on top like this one. did any one had bend problem.is it because all MB are made in china sh**t.this board is in Antec P160 case.
sorry for my bad english.

The temp sensors within the CPU aren't accurate until the temperature approaches Tjmax, so some variation between them, especially at idle, is in my opinion normal. Once it gets up to 50 - 60*C, they should be closer together. If they're not you can always adjust the readings in RealTemp.

Mounting the heatsinks on the boards can cause warping, but when bolted into a case it will be flat so I wouldn't be concerned.

[zsamz_]

Thanks it's a 3849B018.

i got his brother
i should aim for lower volts i guess lol

[xpower]

thanks cryptik.what cooling?how com yours and zsamz batch are same??????

[pcnazz]

World Record , check it out !!! Ctrl + to zoom in .

[zalbard]

It's a completely different board, though...

[Brama]

It's a completely different board, though...

And it costs like a kidney.

[CryptiK]

thanks cryptik.what cooling?how com yours and zsamz batch are same??????

It's on water now. CPU's are made in batches, so there are many in any given batch. Zsamz and I just happened to get ones from the same batch.

World Record , check it out !!! Ctrl + to zoom in http://i470.photobucket.com/albums/r...orldRecord.jpg

Yeah the original X48 Rampage Extreme is such a fun board.

[zsamz_]

Yeah the original X48 Rampage Extreme is such a fun board.

sadly i sold mine today time to move on
it was either the rex or the rexII
the cpu was deciding factor
i7 like a bulldozer

[KILLER_K]

I'm still having issues with my d0 and my kingston hyperx ram. My c0 and this ram got along great but ever since i got this d0 it has been a nightmare. I'm at the point now of just buying some different brand and see what happens. Any body else ran into this problem with the kingston hyper x?

[The Watcher]

I'm still having issues with my d0 and my kingston hyperx ram. My c0 and this ram got along great but ever since i got this d0 it has been a nightmare. I'm at the point now of just buying some different brand and see what happens. Any body else ran into this problem with the kingston hyper x?

that is a strange one, how high do you run it. it's not the ram i think, might be you have the bad proc. have you tried to play the QPI volt.

[CryptiK]

I already posted this in the 4.5 GHz Linx Stable Club thread, but thought I post it here too. The mosfet sinks initially got extremely hot with just my low speed case fans, so I aborted the initial run and positioned a 92mm fan blowing directly onto the PWM sinks. This kept them only warm to the touch, although the capacitors remained quite hot.

Having PWM cooling sorted, I fired up Linx at 4.5 GHz HT On.

Running 20th loop:




Finished:

[The Watcher]

nice one Cryptik. would love to see a 930 performance...

[KILLER_K]

Well if i even touch my ram and try to do it normal it crashes. If i do it under specs it runs fine. Ran memtest86+ and everything passed. At this point i'm about ready to give up on it. As it is odd indeed and how would i know if the cpu is at fault?