X48 Rampage Formula Preview.

[leo27]

@A-Grey
I wouldn't be the least surprised knowing asus to find that there is a problem with the clock skews and high fsb.

I've read a post by mikeyakame in the GTL settings thread where he got his quad q6600 to an FSB of 483 i think it was on bios 701 so with no clock skews on the rampage formula. I'm not sure if his board is unique but i do know he was using a vNB of 1.6-1.7 but it shows that it is possible. I have his bios settings if your interested.

[Alien Grey]

@A-Grey
I wouldn't be the least surprised knowing asus to find that there is a problem with the clock skews and high fsb.

I've read a post by mikeyakame in the GTL settings thread where he got his quad q6600 to an FSB of 483 i think it was on bios 701 so with no clock skews on the rampage formula. I'm not sure if his board is unique but i do know he was using a vNB of 1.6-1.7 but it shows that it is possible. I have his bios settings if your interested.

I know his settings. I tried with changing the NB GTL Voltage Reference to 0.63X and higher FSB termination Voltage but it doesn't make it better.


Here are a few pictures.

This one is with Ai ClockTwister Lighter and DRAM Static Read Disabled. You can compare the bandwidth with the one I posted before with Ai ClockTwister Light and DRAM Static Read Enabled.




This one is with higher NB Voltage running MemTest For Windows. I should say trying to run MemTest for Windows.

[Alien Grey]

I received a reply with a BIOS rom attached. The BIOS rom is called Rampage Extreme 1202 Bios. There's no explanation what to do with it.

Perhaps it's a modded BIOS that can be used on the Rampage Formula.

Would it be as simple as that to use a BIOS rom from the ASUS Rampage Extreme?

What would happen if I try to flash it with this BIOS? Does ASUS EZ Flash II gives you a warning if you try to flash it with a wrong BIOS?


I replyed back asking if this is a mistake but I didn't receive an answer back yet.

[OLDcomer]

Moded ASUS Rampage Extreme bios would be great. That would be a lot of fun.

[Alien Grey]

ROM ID in the file is incompatible with existing BIOS!

Definitely a misstake.

[greg.m]

@SoLoR

This Gigabyte BIOS on my ASUS Rampage Formula is just a joke. I know it would brick my board. I just wanted to say that Gigabyte can make a BIOS that give you a stable high FSB overclock and ASUS can't or won't make one.


@leo27

I'll ask to update the ROM.


@greg.m

Try to get a stable running board with 9 X 470MHz and DDR 1253MHz with tRD6 and let us know how you are doing.



I'll prepare another mail to ASUS and this time it won't be just for my contact at ASUS Tech Support.



There's only one thing left to try for me and that is don't use the CPU and NB Clock Skew. I've got a feeling that this also has something to do with the failure on high FSB.

Im trying it right now...

[greg.m]

Im trying it right now...

No way... to pass memtest with this mem speed itried everything but nothing. dont know if its my system or not but its impossible. Has someone run error free >1200 speeds with rampage and blade or trident modules?

[leo27]

even though the bios they sent you is wrong i'll take it as a sign that at least they seem to be doing something about these problems and that a new bios could be forthcoming, they could have just not replied to you at all. Maybe i'm just an optimist.

[traveler]

@Traveler
Try going into the bios --> boot --> hard disk drives here you can select which hdd is number 1, which is 2 and which is 3. This is just a hunch and may not be what your after at all.

If this doesn't work you could try changing the sata mode from ide to ahci. AHCI enables extra sata features such as NCQ, hot swap etc.
If you have XP follow this guide http://forums.pcper.com/showthread.php?t=444831
If your on vista then follow this http://support.microsoft.com/kb/922976

After you follow the steps in the link you then install intel matrix storage manager and reboot and your good to go. I've already performed this on my rig and it worked fine so i know it works.

The reason changing from IDE to AHCI would fix your problem is that the master/slave naming is from IDE, when you change to AHCI it simply says HDD 1, 2 etc.

Thanks leo.

I think you're right. I think it is a AHCI vs IDE mode thing at work here on my SATA drive configuration. I appreciate the links and will investigate converting to AHCI mode. Will there be any problems running it under XP?



__________________
Asus X48 Rampage Formula / Intel Q9450 2.66 @ 3.2 / 2 x 2GB OCZ Reaper HPC PC2 8500 /
Asus Radeon HD4870X2 Tri-Fan 2GB graphics card / 24" DeLL Ultrasharp 2408WFP monitor /
Western Digital Raptor X Hard Drive / Two Western Digital Caviar RE2 WD5001ABYS 500GB 7200 RPM SATA 3.0Gb/s Hard Drives /
Creative SoundBlaster X-Fi Elite Pro 7.1 / Creative Gigaworks S750 7.1 Speaker System /
Lian Li 343B cube case / PC Power & Cooling Silencer 750 Crossfire Edition /
IOGEAR GCS1782 GCS1782 Dual-Link DVI KVMP switch with 7.1 audio

[leo27]

@traveler
Windows XP will run fine with AHCI just make sure you install the intel matrix storage driver (after following the steps in the guide) and you'll be good to go

[leo27]

I've been having some problems getting my new overclock stable that i posted here
http://www.xtremesystems.org/forums/...postcount=3170
Heres my voltage settings

DRAM Static Read Control [Enabled]
Ai Clock Twister [Stronger]
Ai Transaction Booster [7]

CPU Voltage [1.40000v]
CPU PLL Voltage [1.50v]
North Bridge Voltage [1.39v]
DRAM Voltage [1.84v]
FSB Termination Voltage [1.36v]
South Bridge Voltage [1.05v]
SB 1,5V Voltage [1.50v]
Loadline Calibration [Enabled]
CPU GTL Voltage Reference [0.65x]
NB GTL Voltage Reference [0.63x]
DRAM Controller Voltage REF [AUTO]
DRAM Channel A Voltage REF [AUTO]
DRAM Channel B Voltage REF [AUTO]

CPU Spread Spectrum [Disabled]
PCIE Spred Spectrum [Disabled]
Cpu Clock Skew [Delay 200ps]
NB Clock Skew [Delay 100ps]

CPU Ratio Settings [08.0]
C1E Support [Enabled]
CPU TM Function [Enabled]
Vanderpool Technology [Enabled]
Execute Disabled Bit [Enabled]
Max CPUID Value Limit [Disabled]

These settings are 50 pass LinX stable and 10hrs smallFFT's stable so i know the CPU voltage is fine.
The problem is that when i ran UT2004+P95 Blend it failed at 1024 or 768 size FFT's after around an hour or so.
So i increased FSB termination voltage to 1.38v and got upto 6hrs UT2004+P95 until i stopped it (it didn't error out).

The wierd thing is that now after making this change LinX is no longer lasting more than 5 passes!!!!!!!

I've tried playing with the NB voltage and got no improvement, i"ve set tREF back to 3120T and also got nothing.

So what should i do???

[Alien Grey]

I've been having some problems getting my new overclock stable that i posted here
http://www.xtremesystems.org/forums/...postcount=3170
Heres my voltage settings

DRAM Static Read Control [Enabled]
Ai Clock Twister [Stronger]
Ai Transaction Booster [7]

CPU Voltage [1.40000v]
CPU PLL Voltage [1.50v]
North Bridge Voltage [1.39v]
DRAM Voltage [1.84v]
FSB Termination Voltage [1.36v]
South Bridge Voltage [1.05v]
SB 1,5V Voltage [1.50v]
Loadline Calibration [Enabled]
CPU GTL Voltage Reference [0.65x]
NB GTL Voltage Reference [0.63x]
DRAM Controller Voltage REF [AUTO]
DRAM Channel A Voltage REF [AUTO]
DRAM Channel B Voltage REF [AUTO]

CPU Spread Spectrum [Disabled]
PCIE Spred Spectrum [Disabled]
Cpu Clock Skew [Delay 200ps]
NB Clock Skew [Delay 100ps]

CPU Ratio Settings [08.0]
C1E Support [Enabled]
CPU TM Function [Enabled]
Vanderpool Technology [Enabled]
Execute Disabled Bit [Enabled]
Max CPUID Value Limit [Disabled]

These settings are 50 pass LinX stable and 10hrs smallFFT's stable so i know the CPU voltage is fine.
The problem is that when i ran UT2004+P95 Blend it failed at 1024 or 768 size FFT's after around an hour or so.
So i increased FSB termination voltage to 1.38v and got upto 6hrs UT2004+P95 until i stopped it (it didn't error out).

The wierd thing is that now after making this change LinX is no longer lasting more than 5 passes!!!!!!!

I've tried playing with the NB voltage and got no improvement, i"ve set tREF back to 3120T and also got nothing.

So what should i do???

Why did you change your settings on your stable overclock to find out that you have problems now with changing them?

I don't think it's necessary to use NB GTL Voltage Reference 0.63X with FSB 425MHz. You're using more FSB Termination Voltage to compensate for the lower CPU Voltage but it's most likely causing the failures in LinX.


It isn't the best season of the year in Australia at the moment to do some stress testing on air. The best investment you could do for your rig is buying a watercooling setup.

[Alien Grey]

No way... to pass memtest with this mem speed itried everything but nothing. dont know if its my system or not but its impossible. Have someone run error free >1200 speeds with rampage and blade or trident modules?

Thanks greg.m for testing it with DDR 1253MHz. Let's see what ASUS has to say about it.

[Alien Grey]

even though the bios they sent you is wrong i'll take it as a sign that at least they seem to be doing something about these problems and that a new bios could be forthcoming, they could have just not replied to you at all. Maybe i'm just an optimist.

Yes, they do. The question is if they want to fix it in a new BIOS release.

The board seems to be doing fine with CPU Clock Skew delayed 100ps more than the NB but for some reason it can't stay stable anymore when I use an FSB > 450MHz.

[greg.m]

Yes, they do. The question is if they want to fix it in a new BIOS release.

The board seems to be doing fine with CPU Clock Skew delayed 100ps more than the NB but for some reason it can't stay stable anymore when I use an FSB > 450MHz.

I have exactly the same problem with skews...soo weirdI was stable @ 480 x 9 but i tried some other settings - skews, AICT, volts - and after i went back, itwasnt stable anymore. I need to cmos to find stabilty again @ 480 x 9, everytime the same s..t. Besides the intel speedstep setting was totally disappeared Maybe its time to re-flash bios

[leo27]

@A-Grey
The reason i changed my settings from my stable 24/7 OC 3.4Ghz at 1.4125 (bios) to 1.4000(Bios) was to see if i could get it stable with lower voltage.

I was reading through the GTL thread when a post by mikeyakame made me want to try and see if this was possible to lower my vCPU by getting the GTL's set correctly. He describes how FSB Term is supposed to be 0.02-0.04v higher the vCPU as it acts as a buffer for vCore during overshoot mainly during power on. If it is actually lower than vCore then there is a higher risk you can damage a CPU from too high vCore. It is also basically a limiter for valid BCLK high.

That is the reason my vFSB is so high. I think you are right however that my NB GTL should be 0.67x so i'll test that in the coming days, i don't think it's a CPU vCORE problem since if it was it wouldn't be able to pass 10hrs smallfft's and 50 pass LinX at this new lower vCORE its probably a GTL and FSB Term problem.

[-Acid-]

Just thought i would add this to the mix

[Alien Grey]

Just thought i would add this to the mix

How stable is it with FSB 500MHz?

[leo27]

@-Acid-
Bios settings would be nice

[leo27]

@A-Grey
Any news on a possible new BIOS??

[-Acid-]

hi leo27 sadly it isnt stable at all computer is appart atm getting rehoused so will post up settings when its rebuilt

hopefully today

[Alien Grey]

@A-Grey
Any news on a possible new BIOS??

No, I didn't get a reply back on my last mail to ASUS.

hi leo27 sadly it isnt stable at all computer is appart atm getting rehoused so will post up settings when its rebuilt

hopefully today

A Quad and FSB 500MHz is almost impossible to have stable on this board or any other ASUS board. The BIOS isn't optimized for high FSB and Quads.

I did some LinX testing at 9 X 500MHz and can't keep it stable for more than 2 passes. Joining the Intel 4.5GHz LinX Stable Club isn't going to be for tomorrow.


It's time that ASUS realizes that they should do something about that the only ASUS board in the Club is an ASUS Rampage Extreme with a Dual. Don't they use Quads at ASUS.

[OLDcomer]

Shame on Asus. I have hard time benching over 509fsb. A new bios with increased overclocking potential for quads, precise clock skew adjustments would do the job.

[Alien Grey]

Shame on Asus. I have hard time benching over 509fsb. A new bios with increased overclocking potential for quads, precise clock skew adjustments would do the job.

You're right. The CPU and NB Clock Skew should have increments of 50ps. I don't know if this is possible with ASUS boards but if it's possible that would help a lot.

Ai Clock Twister Light and Lighter could be used for optimizing the BIOS for Quads and high FSB. Less is more as you could see in my EVEREST results but it's probably not enough.

Maybe with need Ai Clock Twister Lightest.

[OLDcomer]

As a matter of fact, I use Strong Twister only and have slightly better Read and Latency but at a price of higher NB voltage.