Live MB OC Report :: ASUS Rampage Extreme

[vega22]

Hi Vega22, any chance of a picture ?. I have the same H/S and would like to explore after market N/B coolers. Did you have to remove the complete heatpipe assembly to fit the HR-05 or does it sit in place of the fusion block ?.

Sorry...now i have only the bad cam in my mobile phone. Of course i have removed the entire stock heatsink. Then i have installed the Thermalright HR-09 Type 2 in the mosfets areas, the HR-05 SLI/IFX in the NB w/8-1.5" fan, the Enzotech SLF-1 in the SB (Dremel modded) and Enzotech BMR-C1 (and anothers ram heatsinks) for the others minor mosfets/caps around the board (Dremel modded here and there).

[KanGaXx]

Y KNOW but i tryed them on single channel and they work!!!!
if they dont work on dual channel i think is a problem of motherboard of a setting or something similar!!
i tryed cas9, 10 and always the same... there somethink i need to set!
i tryed vNB 1,6 and greater but nothing.. i tryed 2,2vRAM and nothing...
help me pls
can someone post me some setting for 1000mhz run ?
DRAM Skew ?
NB ref ??

Look at the Home Page from CSX.
They work with a X48 Chip at 1866 MHz.
2000 MHz was on a Nforce Chip.

[Bullshooter]

The traffic on the NB was to high for 2000 MHz.
You must wait for a new Bios.
We have call CSX (Compustoxx) and they said the RAM max works at 1700 MHz on the Rampage Extreme. (The 2x2GB Kit)
You can go higher, its work prime stable, because play Crysis and you have freeze or blue Screens.

[KanGaXx]

i noticed that if i put DRAM skew CTL , CLK, and the other on channel B : advance 100ps i have better stability... i can run OCCT more time about 5-6minutes
if i leave those on AUTO i can run 30-60sec.....

can be the solution ?? someone can explain me more?

[3oh6]

D9GTR and this motherboard don't work well together, my G.Skill D9GTR kit runs 1100MHz 32M stable with reasonable voltage on 790i...can't even get Windows to load on the RE with that kit at 1000MHz 8-7-6. i would suggest you spend less time asking for help here and more time testing the system because you would perhaps be the first to get 1000MHz 8-7-6 with D9GTR to run anything. also, getting them 100% stable at 1000MHz isn't going to be easy on any motherboard. on top of that OCCT failing is usually a sign of something else, it has never been even a reasonable test for memory for me. stick with dual 32M SPi for stability testing then work your way up to Prime Blend once you get that stable.

[KanGaXx]

ok thanks a lot
i appreace what you say thank too much.
i am spending a lot of time testing i am only talking to you for try to make it best as possible

[MnM]

I dont seem to have any problems with my CSX Diablo memory at all.

OK here is where I am so far:

Rampage Extreme 501 bios
4 x 2 Diablo D9GTR based

[ShoNuff]

I dont seem to have any problems with my CSX Diablo memory at all.

OK here is where I am so far:

Rampage Extreme 501 bios
4 x 2 Diablo D9GTR based

Wow! Impressive especially since you are using a 4x2GB config. What volts for the ram? Looking at your cpu temps it looks like you are on air correct?

It looks like the 501 bios is a keeper!

[Mix]

Hi

Look the spd of one kit csx diablo 2x2gb gtr 2000



How can i change this spd????

[JaVa_Az]

First off i am going to refer to ROG Speeding HDD as Speedy ( although from the tests slow poke might be a better name for it) Test Drives are 150GB WD Raptors

First up here we have the linear read from everest on the Speedy controller. As we can see it came up with a disappointing 96MB/s for a maximum. To save time i just did the first part of the drive. I wanted to focus more on the File system benchmarks which are coming up in a bit. I believe these will show a more real world result since i would say most of us have are drives formatted to the NTFS file system.

Now we have the Intel ICH9R Raid 0 with a stripe size of 32KB. From playing around with stripe sizes the 32KB yielded the best performance on my hard drives. As we can see from the test results a 175MB/s maximum linear write Puts a serious whooping on the speedy. Very close to a 100% gain!!!!

But these tests are raw writes Lets see how the Two controllers Fair in a more real world test With Si soft Sandra's File System Benchmark. A quick note you can perform this test without destroying data So anyone can do it and you don't need your OS on a separate drive. First Up we have the Speedy results.This shows a Drive index of 102MB/s . This test combines read, write and latency to come up with the score.

Now we have the Intel ICH9R In raid 0 32MB stripe. This shows a score of 127MB/s for a drive index. So not Quite the Romping it gave the Speedy in the linear write tests But If my math is correct which it very well Might not be, I have the Intel Up by Just shy of 20% over the speedy


now lets look at the linear Write tests of One Single WD 150GB raptor on the two different controllers. We have speedy first

Now we have One raptor on the Intel

Notice i have a couple bad spots in one of my drives. Teach me not to buy OEM ever again. But these test are about the same. Not much of a difference . Not for sure but i believe that the Speedy ports are some how combined into one channel. But this is just a theory , cant find any tech specs on it.
So sitting there with my two drives non raided on the Intel controller I decided to get crazy and converted them to dynamic disks in windows disk management then i proceeded to make a software raid stripe. I have read quite a few forum posts where people say that Windows software raid is not good that hardware raid is a lot faster. So lets take a look at the windows software raid 0 results. Wait can this be the software Raid 0 shows a 127.95MB/s index beating the Intel ICH9R hardware raid that had a drive index of 127.42MB/s. Wow a 0.5 MB/s victory for the Windows software Raid 0. To bad you cannot install windows on it. Just to be sure i ran that test 3 times came out the same. ( Note there was no way to do any other tests on the software raid because all the benchmark programs saw the raid as two separate drives. So File system was the only choice i had )[/LEFT]


Now this next test i didn't think would be possible i turned the speedy to normal mode with both drives in the speedy ports. The Rampage extreme manual says this normal mode only supports one drive. Well windows saw both drives installed the drivers. So I converted to dynamic disks slapped a software raid 0 on them and ran the file system benchmark. Once again software Raid O with a win. With a drive index of 111MB/s beating out the Speedy controller by 9MB/s . But still slower then the Intel ICH9R tests
[/LEFT]

From my first findings i said that windows booted up quicker with just the speedy enabled which it did. So i decided to give the Intel the same opportunity and hooked my external dvd Esata drive up via USB. Disabled Marvell completely and my windows boot up time is even quicker then it was on the speedy controller. So either way you look at it if windows has to load both the Marvell and ICH9R drivers it will just take quite a bit longer. Anyways i know i had more to say about this but i have this crazy mod i am dieing to work on today ..So I'm Off till next time

[DeathWalking]

I dont seem to have any problems with my CSX Diablo memory at all.

OMFG. You have my exact setup. Please post complete list of settings!

[3oh6]

for those of you who aren't LCC whores like myself, Gautam showed me how to make the MCW30 fit directly on
the X48 IHS...



i am using some crappy cooler master generic heat sink stuck to the SB with Sekesui thermal tape and the rest
of the heat sinks are MC14s...with a couple cut in half for the small MOSFETs around the NB. the result...

600x6 5:8 960.0 7-6-5 Performance Level 8 w/1.75vNB...NB temps never crossed 34C during this run

click for full size including memset & voltages

[jugeen]

i like your RAM size, Jody!

[3oh6]

i like your RAM size, Jody!

haha, they are a custom kit i did at work. you know, took off a couple IC's and replaced them

i didn't even notice that, but i use an edited CPU-Z ini file so it opens faster and it must be picking up my
maxmem value now

[zlojack]

I've just finished setting up the board in my new case. Still not on water yet, since I'm waiting for more components, but up and running at least. Funny enough, it booted into my old Windows install on the two RAID0 disks, but it's brutal and missing drivers.

I'll be back after a reinstall.

Since I'm on stock cooling on my CPU, I won't be doing much overclocking yet.

[tekjunkie]

I have a very interesting problem. With my Gskills and my QX9650, everything on Auto, Memtest 2.01 fails miserably, SPI 32M sometime passes and sometime fails, Intel burn test passes. It also passes prime for more than 7 hours before one the cores crap out, but that's understantable since @ 400 FSB 1:2, it might need some GTL tweaking. I think it's time for some SSD testing with my core V2 on the ICH9R in single disk, 4 of them in RAID 0 on the ICH9R and the same on the Areca.

[Mezzro]

Java, excellent work with the side-by-side native RAID controller benchmarks. Thanks!

[MnM]

Wow! Impressive especially since you are using a 4x2GB config. What volts for the ram? Looking at your cpu temps it looks like you are on air correct?

It looks like the 501 bios is a keeper!

A few pages back Revogamer posted his config, so I tried that as well. Before this I was stable at 400FSB x 10.
Still tweaking to see what can be lowered but the below config is 100% stable.

CPU is on water the rest is stock cooling with washer mod. This is in a closed case, not on bench or sides open.

Extreme Tweaker Menu

CPU FSB - Manual
CPU Level Up - Auto
FSB: 430
-cpu configuration
-cpu clock skew - Auto
-nb clock skew - Auto
FSB Strap to North Bridge - 333
DRAM FREQUENCY - 1721
DRAM Command Rate - Auto
DRAM Timing Control - Manual
*1st information : 8-7-6-15-Auto-Auto-Auto-Auto
*2nd informaion : all AUTO
*3rd information : all AUTO
DRAM Static Read Control - Auto
DRAM Dynamic Write Control - Auto
-DRAM Skew Control
Ai Clock Twister - Auto
Ai Transaction Booster - Auto
*C/P : A1 A2
*LVL : 07 07

EPU II Phase Control - Auto
-----------------------------------------------------------------
*** Please Key in Numbers or select voltage!***
CPU Voltage - 1.462
Load-Line Calibration - Enabled
CPU PLL Voltage - 1.709
FSB Termination Voltage - 1.45
CPU GTLVref (o) - +40
CPU GTLVref (1) - +50
CPU GTLVref (2) - +50
CPU GTLVref (3) - +40
NB GTLVref - +30
North Bridge Voltage - 1.66997
DRAM Voltage -1.96156
NB DDRVref - Auto
DDR3 ChannelA Vref - Auto
DDR3 ChannelB Vref - Auto
South Bridge 1.5 Voltage - 1.6
South Bridge 1.05 Voltage - 1.08
**********************************************
CPU Spread Spectrum - Disabled
PCIE Spread Spectrum - Disabled

[ShoNuff]

MnM thanks for the info. It makes me want to pop my Quad back in and see what it can do on the RE. On my 790i it was OCCT stable at 450x9 (QX9650) and I could bench up to 4.2GHz but it was far from stable.

I sent a PM to Kasparz regarding the CSX Diablo's and he said he is still looking for a distributor in the US. Your results are also impressive and I wish my ram could do the same.

[jugeen]

haha, they are a custom kit i did at work. you know, took off a couple IC's and replaced them

i didn't even notice that, but i use an edited CPU-Z ini file so it opens faster and it must be picking up my
maxmem value now

i also have a custom kit with dual or half spd table

http://valid.x86-secret.com/show_oc.php?id=419143

[dread77]

I just got my hands on this sweet piece of hardware and I can say that, besides those crazy FSBs, there is nothing that Evolution or Blackops can't do better. Like...

- Stability
- Temps... This board is hotter than a bbq grill.
- Voltages. Once again Asus brought a high-end mobo that is overvoliting like hell.

Though, extraordinary FSBs + X48 makes this board the best C2D bencher around...

[Viss]

Very nice Dread. You use 400 strap? There's a problem with 3/5 divider on X48 and 400 strap. Its actually 2/3 so mem wil be 900mhz.

[over@locker886]

for those of you who aren't LCC whores like myself, Gautam showed me how to make the MCW30 fit directly on
the X48 IHS...



i am using some crappy cooler master generic heat sink stuck to the SB with Sekesui thermal tape and the rest
of the heat sinks are MC14s...with a couple cut in half for the small MOSFETs around the NB. the result...

600x6 5:8 960.0 7-6-5 Performance Level 8 w/1.75vNB...NB temps never crossed 34C during this run

click for full size including memset & voltages

Very nice work my friend! Your temp is insane good..

[Hazzan]

with diveder 3:5 still have bug on CPU Z......

[eternal_fantasy]

Did you play with skew settings on this run? wondering if I should bother to get over 600FSB...