X48 Rampage Formula Preview.

[Alien Grey]

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.

I was using the 400MHz strap. With the 333MHz strap I don't have that problem at least not noticeable in EVEREST. But the problem is still there because it doesn't matter what you select for Ai Clock Twister it still fails at high FSB.

We could say that it's because of the X48 chipset but that doesn't explain why all other ASUS boards with newer chipsets also fail at high FSB with Quads.

[leo27]

Just a thought but have you tried setting PCIE Frequency to 101-103mhz or 110-114mhz???

I've read that it can increase stability but don't ask me how, these ranges were determined by mikeyakame and gave him extra stability on the Rampage Formula

EDIT:The reason it adds sability is apparently when set to 100 and you OC the FSB it raises the PCIE frequency when it isn't supposed to. Setting it to 101 or higher sets the value permantly and thus allowing for a bit more stability. You may have to slightly increase the SB voltage a little though

[Alien Grey]

Just a thought but have you tried setting PCIE Frequency to 101-103mhz or 110-114mhz???

I've read that it can increase stability but don't ask me how, these ranges were determined by mikeyakame and gave him extra stability on the Rampage Formula

EDIT:The reason it adds sability is apparently when set to 100 and you OC the FSB it raises the PCIE frequency when it isn't supposed to. Setting it to 101 or higher sets the value permantly and thus allowing for a bit more stability. You may have to slightly increase the SB voltage a little though

I've tried that but it doesn't help.

I've spend a few hours again trying to get it stable at 9 X 500MHz. This time with 1:1 divider but it's still a no go.

[-Acid-]

A-Grey how long are you getting before crash ?

[Alien Grey]

A-Grey how long are you getting before crash ?

The longest stable run was 4 passes LinX. Not enough to join the Intel 4.5GHz LinX Stable Club.


I wrote a second mail to ASUS asking them to optimize Ai Clock Twister Light and Lighter for high FSB overclocks. I also asked again if it would be possible to have 50ps clock skews and to add tREF and update JMicron.

My Intel Q9650 and my G.SKILL F2-9600CL5D-4GBPI memory are begging for a better BIOS.

[SoLoR]

and update JMicron.

you can update this on your own, if you have any spefic problem i can give you my own bios with 1.07.14 jmicron option rom... but ye, its just wrong that latest bios for rampage formula is still from april, even tho im sure in this time at least 5 people where asking them for random things to add ... also dont think i ever saw asus updating their option roms... for god sake if i rememeber correctly x58 (ich10r) board is still using same intel 8.0.0.x option rom that was introduced in rampage extreme....

[leo27]

@SoLoR
Would you mind posting Bios 902 with the updated JMicron Rom

[SoLoR]

@SoLoR
Would you mind posting Bios 902 with the updated JMicron Rom

sure

http://cid-afe12d66a5ec12fa.skydrive...c/0902ASLI.ROM

this is what im using, so bios is tested and is working fine for me, but most important shouldnt cause any badflashes with no way to recover.

this has:
intel option rom 8.9.1.1002 (im using raid0, works fine, ahci mode also works)
jmicron 1.07.14 (only have 1 ide dvd rom connected, but it works)
asus 2.1 slic
asus sli cert (in case you want to run sli with moded hal.dll/nvidia drivers)

[leo27]

@SoLoR
Thanks for the Bios

I've been doing some testing with PCIE frequency and have some interesting results
This is testing with my 3.4Ghz OC with 1.4000v in Bios (1.384 real), 50 pass LinX stable and 10hrs Small FFT's which fails UT2004+P95 Blend after 10min with PCIE at 100mhz

And here is the HDD performance


Now rasing the PCIE to 103Mhz with the same slightly unstable settings allows it to run UT2004+P95 for 3hrs until error
The HDD performance is here


As you can see it's slighty higher.

The failures in P95+UT2004 in both cases are occuring in LargeFFT's.
I think it an odd coincidence that both mikeyakame and myself have found benefit's by raising the PCIE Frequency and that we both run Q6600's.
Maybe it's something to do with 65nm quads or the Q6600 itself that likes higher than normal PCIE Freq??

This leads me to believe that it is possible to find a setting that will stabilise this OC as i know it isn't vCPU (since it is LinX and SmallFFT stable) or vNB related it has to be GTL's or Skew related (either NB CPU or DRAM)

[phillai]

Can someone direct me in the right direction. I just got some new speakers but I cannot get the sound working via optical.

I sussed out that it might be something to do with the SoundMAX drivers.

Reading some posts it appears that the SoundMAX drivers on ASUS own website are not the latest ones...

Where can I get the latest SoundMAX drivers from, for Windows 7 64 bit OS?

Many thanks to all replies in advance.

[leo27]

the windows 7 drivers on asus' site are the latest for soundmax

[SoLoR]

Can someone direct me in the right direction. I just got some new speakers but I cannot get the sound working via optical.

I sussed out that it might be something to do with the SoundMAX drivers.

Reading some posts it appears that the SoundMAX drivers on ASUS own website are not the latest ones...

Where can I get the latest SoundMAX drivers from, for Windows 7 64 bit OS?

Many thanks to all replies in advance.

feel free to try this ones:

http://cid-afe12d66a5ec12fa.skydrive...600.custom.zip

they are some intel soundmax drivers, with deviceids from asus drivers added by me. However im not using soundmax (i have x-fi), so i didnt try them my self. But feel free to try them and report if they are working at least

[Alien Grey]

I think it an odd coincidence that both mikeyakame and myself have found benefit's by raising the PCIE Frequency and that we both run Q6600's.
Maybe it's something to do with 65nm quads or the Q6600 itself that likes higher than normal PCIE Freq??

This leads me to believe that it is possible to find a setting that will stabilise this OC as i know it isn't vCPU (since it is LinX and SmallFFT stable) or vNB related it has to be GTL's or Skew related (either NB CPU or DRAM)

Higher PCIe Frequency works for all of us but there isn't much performance gain and that's why I always leave it at 100MHz.

Bump the memory and/or NB Voltage 1 step in the BIOS. It's possible that the voltage fluctuates a little bit and drops to low.

[leo27]

@A-Grey
The point about PCIE Frequency wasn't so much the increase in speed but the unexpected increase in P95+UT2004 stabilty.
The NB is completly stable at 1.39v and i have it set to 1.41v. As for vDRAM i haven't tried raising that up a step extra from where it is stable so i'll try it.

I think that it's FSB termination thats fluctiating. I noticed in the Hardware monitor section that it fluctuating between 1.36 and 1.344 when set to 1.30 but i'm not sure it was happening under load or not so i'll have to check that. At any other setting this doesn't fluctuate, at 1.32 it is a steady 1.38.

This means i'll have to re-tet the GTL's however i think at 1.32 and CPU 0.67x NB 0.67x it should be good. The CPU seemed to like 0.65 with vFSB 1.42 real (LinX and P95 SmallFFT's stable) and the NB liked 0.67x with vFSB 1.44 (P95+UT2004 stable but not LinX stable). At 1.32 bios(1.38 real) i would get 1.38x0.67 for CPU and NB which works out fairly close to the mark for both CPU and NB reference voltage.

[Alien Grey]

I've updated the OP on the ASUS Rampage Formula forum.

Check it out and let me know if I've missed something that should be fixed or added to the BIOS. Especially my post with evidence.

For those of you that are members on the ASUS Rampage forum, you can reply in the thread. For those who aren't members, now is the time to take a few minutes of your time and make you a member so you can reply in the thread with what you think should be done with the BIOS.


After all I've done to help their ASUS Rampage Formula users on the forum to get the best out of their hardware or try to solve their problems, they think that they can just ignore me.

Never did I ask anything in return. I did it all for free and was really happy that I was a member of the ASUS Rampage Formula community. All I'm asking is to fix the problems in the BIOS and add some features that should have been there from the day I fired up that board. If that is to much asked than I don't know why I still keep on doing it and should probably dump this board and follow my friends.


Here's the link: http://vip.asus.com/forum/view.aspx?...Language=en-us


Keep in mind that this is my last call to ASUS and that I always give them the link to the this thread and the one on the ASUS Rampage Formula forum. So if you still care to see a better BIOS on the ASUS Rampage Formula I would appreciate if you take a few minutes time to reply in the thread and write down what you think about it.

After all I did to see improvement and when that isn't possible anymore than I just have to leave the ASUS scene.

[leo27]

@A-Grey
I have registered on the ASUS forum and tried to reply to your post however when i click submit on the reply it doesn't post. I'll try again later though. I think i speak for everyone in this thread when i say we appreciate your efforts.

Heres what i was trying to post on the asus forum

There is definitely a big skew problem with this board

This is a bug that should be addressed and should certainly warrant a new bios as it completely un-stabilises the board. Adding 50ps skew adjustments would also go a long way to achieving stability with this board (although this might not happen due to hardware limitations).

As for AI Clock Twister it is unacceptable that this should behave the way it does.
Setting to Stronger should provide the most bandwidth by tightening timings whilst Lighter should provide the least bandwidth by loosening timings (with strong moderate and light providing intermediate steps).
This is another blatant bug which interferes with the function of the board and as such needs to be rectified.

This board was advertised to support DDR2 1200MHz, in order to do this and comply with JEDEC specs we need tREF adjustment as a tREF of 3120t is simply to low at 1200Mhz. Manual tREF adjustment would be a welcome addition to the BIOS and should include an auto setting which sets tREF according to JEDEC specs depending on RAM frequency.
This would greatly increase the ability to achieve the stability we were lead to believe we would achieve through the purchase of this board and as such we all paid a premium for it.

Also it is unacceptable that ASUS haven't updated the JMicron and Intel option ROM's. All other motherboard manufactures frequently update these and it is time for ASUS to follow suit. These updates fix performance bugs and issues relating to the Intel RAID controller and the JMicron PATA controller and we would get these if ASUS just spent about 20min to incorporate the updated ROM's.

We all paid a premium price for this board expecting a high performance X48 chipset that could through everything we could at it with stability and customer service congruent with our expectation of the Republic of Gamers Line.

If ASUS want any of us to purchase another board from them then our concerns need to be heeded and fixed. These are after all bugs that affect the stability of the board.

I can tell you that if all these problems we have found fall on deaf ears once again then this will be my first and last ASUS motherboard purchase, I will go to which ever company listen to their customers about bugs and stability problems.

[Alien Grey]

@A-Grey
I have registered on the ASUS forum and tried to reply to your post however when i click submit on the reply it doesn't post. I'll try again later though. I think i speak for everyone in this thread when i say we appreciate your efforts.

I've posted it for you in the thread. Thanks for the appreciation.

If anyone else has something that he wants to post in the thread and has difficulties to do it, just post it here and I copy and qoute it in the thread. ASUS also has problems with their forum that they can't seem to solve.

[SoLoR]

Also it is unacceptable that ASUS haven't updated the JMicron and Intel option ROM's. All other motherboard manufactures frequently update these and it is time for ASUS to follow suit. These updates fix performance bugs and issues relating to the Intel RAID controller and the JMicron PATA controller and we would get these if ASUS just spent about 20min to incorporate the updated ROM's.

Actually it takes them 1 min (bigest consume of time is finding them, however since they are company that produces mobos, dont think they can have problems on that department)... 20min is more like if they want to update them on ALL motherboards. But ye ASUS is NEVER updating option roms on ANY of their boards, no idea why, when intel for example is updating them constnatly (figures, their chipset, their mobos, their roms) and lot of other manufactures less frequently but they still do it 2-3 times/year. However officaly only 7.x option roms supports ICH9R so dont expect mirricles on this department, 8.x are still working (figures ICH10R is basicaly downsized ICH9R), however 9.x seems to not work at this moment, even intel X58 mobos are using 8.9.x...

I actualy gave up on asus few months ago after i sended them 2 mails about tref... Also it seems i got in to my CPU max stable frequency anyway with 8x451 since it doesnt seems i could make it stable no matter what i did (in normal limits), so i stoped complaining that much, however i probably wont buy another asus board and i was big fan of asus until now. However they have great RMA policy... 2 months ago my old 9800gx2 burned out, i got gtx295 in replacement It took then month+ but still!

If Asus wants then can feed me with newest hardware and ill be taking care of drivers/bios updates for that! Im sure in that case people will be tired of "changes" every week

[Alien Grey]

However they have great RMA policy... 2 months ago my old 9800gx2 burned out, i got gtx295 in replacement It took then month+ but still!

WTF they gave you a GTX295 for your dead 980GX2 but can't give us a decent BIOS for our Rampage Formula.

Their RMA policy is definitely a lot better than EVGA. They couldn't replace my EVGA GTX 285 Hydro Copper that was DOA.

[leo27]

@A-Grey
Thanks for posting my thoughts on the asus forum, could you put in the post that my nickname on the asus forum is leo27 as well so they know that i am registered(i just couldn't post).
Let's hope we get some results

[Alien Grey]

@A-Grey
Thanks for posting my thoughts on the asus forum, could you put in the post that my nickname on the asus forum is leo27 as well so they know that i am registered(i just couldn't post).
Let's hope we get some results

Try this. Login on the forum and go to the ASUS Rampage Formula main page. Enter the thread you want to reply in and try to post your message.

It's an annoying bug with the forum that you can't post in a thread that you opened if you didn't login first.

[leo27]

@A-Grey
Thanks but i've already tried that and no joy, it's like ASUS don't want me to post

After more testing i've come to the conclusion i will never be able to get stable at a CPU voltage of 1.4000 bios (1.384 real). I can make the CPU perfectly stable, however the mobo fails to keep it stable during UT2004+Blend and if i get it UT2004+blend stable then the CPU fails.

The issue derives from the miserable GTL adjustments avaliable, and i'm not sure but i think the skews may be contributing aswell since if the skew is CPU -100ps and NB normal i get a BSOD whilst on CPU normal NB normal it goes for about 30 passes. As i understand it skews should increase stability not decrease it.

The reason i say that it could be a GTL problem is that with vTT at 1.36 i'm CPU stable (LinX and SmallFFT's) at CPU GTL 0.65 NB GTL 0.63 but this will fail UT2004+blend after some time. If i increase FSB Term to 1.38 i am now UT2004+blend stable for however long i test but the CPU will now fail LinX.

If i change NB GTL to 0.67 or 0.63 with vTT at 1.36 then i fail LinX also.


http://edgeofstability.com/articles/.../gtl/gtl6.html

The above is an image showing what happens regarding vtt and gtl reference voltage where the pink trace shows what happens when the input is held high and the blue trace is when the input is held low. As you can see if we had control over Vih(min) and Vil(max) we could have a much better overclock and it is exactly this which is implemented on the p45 based boards and why boards such as the ga-ep5-ud3p series can reach fsb's of 550's which is unheard of for a rampage formula.

I'm pretty sure that i'm at a point where raising the gtl reference causes Vih(min) to hit the pink trace, and lowering the gtl causes Vil(max) to hit the blue trace, but any lower vtt also causes Vil(max) to hit the blue trace even at 0.67x.

If i up the CPU voltage to 1.4125 in bios then i can force it to be stable (in effect lowering the peak of the blue trace or improving the signal quality) with FSB term 1.36 CPU GTL 0.65 NB GTL 0.67, however with a board with precise GTL settings and finer Skew adjustments i probably could get it stable with lower vCPU.

Basically i need a CPU reference voltage of around 0.923 with a GTL ratio of 0.65x, a NB reference of 0.9072 with a GTL ratio of 0.67x whilst simultaneously having an FSB Termination of no lower than 1.36 bios (1.42 real) together with an appropriate CPU and NB clock skew to keep it all stable.

Thus if i had these avaliable and skews which actually worked i could probably reach stability. I know this is quite technical but i am an engineering student so my appologies.

Now i know there is no way to get the added gtl settings in a new bios however proper CPU and NB skew adjustments which work correctly would allow for an increase in signal stability and thus a higher chance of oc sucess as they also have the effect of creating a more stable signal (mikeyakame had an accurate discription in this thread somewhere).

[Alien Grey]

I'm not an engineer but I know what you 're saying. I've got the same poblem with my Q9650. I need more CPU Voltage to get it Prime95 Blend stable for more than 6 hours than I need for 100 passes LinX.

It isn't right but there's not much we can do about it. As long as we can get it stable ASUS is never going to give us better GTL adjustment in the BIOS.


Anyway our complaints and requests arrived at the ASUS RD team. I hope that they can fix and add everything that we've been asking.

It took a little longer to get a reply back from ASUS but it isn't because they didn't want to.

[sirheck]

Ok guys...should i update my BIOS?

Im currently running the 0403 BIOS.

http://support.asus.com/download/dow...Language=en-us

[leo27]

I would say move to Bios 902, there is no real reason why not