X48 Rampage Formula Preview.

[JohnZS]

Perhaps it might be worth e-mailing ASUS support. Because if an ASUS eePC works with it then they could test what is making it work on one but not the other.
John

[Counter CS]

I got an MP3 frome Creative buyed last week, its Mosaic EZ100 16GB, same thing happens, not detected.
On EeePC 1005HA its all fine.

[Alien Grey]

I personally believe this is a problem with a lot of the 775 ASUS boards.
Difficult (or perhaps impossible to prove), but I swear some settings which are [AUTO] do not remain consistent after a COLD boot.

I.e. you input settings, stability test, fail, reboot, change settings, stability test, fail, reboot, change settings some more, stability test... and then everything checks out fine.
20x Linx
Hiper Pi
3dmark Vantage Loop.
3dmark 06 loop.

And then play a game for a bit, etc. Turn off PC at the end of the day.

Boot the next day, instant BSOD System Service exception or Stop Error.

POWER OFF PC (i.e NOT reboot), then power on and things are fine.
However if you reboot (reset or a quick power off and on) BSODs continue

I have had this happen on two ASUS boards, P5E3 and Rampage Extreme. It appears you are having a similar issue too.

Sadly nobody can prove this as it is so difficult to pin down.

John

The BIOS calculates some of the timings according to the BIOS settings every time it boots up. When your DRAM Clock Fine Delays aren't the same on every boot or reboot than you know that your overclock is unstable. A difference of more than 1T is bad.

The calculation fails to easy and that's why it's so hard to find a stable overclock on an ASUS board.

I think i just have to live with that until my next build Ill probably go for a msi or gigabut board this time...and i hope that Alien Grey will follow too ...

Anything but ASUS.

I guess it's time to save some money for a new build.

I got an MP3 frome Creative buyed last week, its Mosaic EZ100 16GB, same thing happens, not detected.
On EeePC 1005HA its all fine.

Maybe there's not enough voltage on the USB connectors.

[Senator]

Have you guys tried the Intel BIOS Implementation Test Suite yet?

http://www.xtremesystems.org/forums/...d.php?t=267289

Do at test run, maybe AlienGrey can make some sense

[Alien Grey]

Have you guys tried the Intel BIOS Implementation Test Suite yet?

http://www.xtremesystems.org/forums/...d.php?t=267289

Do at test run, maybe AlienGrey can make some sense

Running MSR consistency test suite

(MSR 0x0 consistent): FAIL
(MSR 0x2a consistent): FAIL
(MSR 0x39 consistent): FAIL
(MSR 0x118 consistent): FAIL
(MSR 0x19e consistent): FAIL
(MSR 0x400 consistent): FAIL

Summary: 4687 passed, 6 failed


SMI Frequency and latency

Summary of impact: observed maximum latency = 511us
Risk of OS timeouts exists with this observed latency (ideal is < 150us)


It's very bad but it could be worse. These test were done at 4.05GHz. The maximum latency is +600us at stock speed and BIOS Defaults.

[JohnZS]

Running MSR consistency test suite

(MSR 0x0 consistent): FAIL
(MSR 0x2a consistent): FAIL
(MSR 0x39 consistent): FAIL
(MSR 0x118 consistent): FAIL
(MSR 0x19e consistent): FAIL
(MSR 0x400 consistent): FAIL

Summary: 4687 passed, 6 failed


SMI Frequency and latency

Summary of impact: observed maximum latency = 511us
Risk of OS timeouts exists with this observed latency (ideal is < 150us)


It's very bad but it could be worse. These test were done at 4.05GHz. The maximum latency is +600us at stock speed and BIOS Defaults.

I have the same failings on my Rampage Extreme!
See here for more details

From what Zailbird and nn_step are saying ASUS could quite easily code a fix for the MSR failings and the option of HPET ON/OFF would help considerably with the latency.

John

[Leeghoofd]

Till now tried it on Asus and MSI boards ( AM3 790FX and 890FX and a bunch of H/P67) : all fail

Edit : The H67 M Evo passed all tests

[Alien Grey]

I have the same failings on my Rampage Extreme!
See here for more details

From what Zailbird and nn_step are saying ASUS could quite easily code a fix for the MSR failings and the option of HPET ON/OFF would help considerably with the latency.

John

I also tested the latest official BIOS 1001. It has the same 6 MSR fails and the SMI Frequency and latency is 676us with BIOS default.

Now we know that the 6 MSR fails and very bad SMI frequency and latency aren't because of SoLoR's MOD BIOS 1002.

[Counter CS]

new version

"bits-329"

Posted Sat Feb 26 22:04:50 2011

Fix a bug in the SMI Latency test which caused GPF reset when the SMI Count MSR is unsupported by the processor. Be sure SMI Count MSR read is only read via SMP function to ensure that GPF handling is performed.

[Counter CS]

MSR consistency test suite:

( MSR 0x39 consistent ) FAIL
( MSR 0x14e consistent ) FAIL
( MSR 0x1dd consistent ) FAIL
( MSR 0x415 consistent ) FAIL

Summary: 4851 passed, 4 failed...

Detect SMI frequency and latency:

Starting pass 1. Calibrating the TSC.
Starting pass 2. Wait here, i will be back in 15 seconds
0 < t <= 1us; average = 70ns; count = 211623293
Times between first few observations: 126ns 105ns 66ns 66ns 63ns
100us < t <= 1ms; average = 598us; count = 57
Times between first few observations: 256ms 513ms 256ms 256ms 256ms
1ms < t <= 10ms; average = 1866us; count = 1
Times between first few observations:
Summary of impact: observed maximum latency = 1866us
risk of OS timeouts exists with this obsrved latency ( ideal < 150us ).

Test round-trip latency via MWAIT

elapsed time = 96ms with 682 iterations/ms

[Alien Grey]

MSR consistency test suite:

( MSR 0x39 consistent ) FAIL
( MSR 0x14e consistent ) FAIL
( MSR 0x1dd consistent ) FAIL
( MSR 0x415 consistent ) FAIL

Summary: 4851 passed, 4 failed...

Detect SMI frequency and latency:

Starting pass 1. Calibrating the TSC.
Starting pass 2. Wait here, i will be back in 15 seconds
0 < t <= 1us; average = 70ns; count = 211623293
Times between first few observations: 126ns 105ns 66ns 66ns 63ns
100us < t <= 1ms; average = 598us; count = 57
Times between first few observations: 256ms 513ms 256ms 256ms 256ms
1ms < t <= 10ms; average = 1866us; count = 1
Times between first few observations:
Summary of impact: observed maximum latency = 1866us
risk of OS timeouts exists with this obsrved latency ( ideal < 150us ).

Test round-trip latency via MWAIT

elapsed time = 96ms with 682 iterations/ms

The high latency doesn't look very good.

Maybe you should try to do the Rampage MOD and flash it with SoLoR's MOD BIOS 1002.

[Alien Grey]

I decided to give it an other try.

Dear XXX,


The Intel BIOS Implementation Test Suite gives 6 SMR fails with latest
ASUS Rampage Formula BIOS 1001 and to high SMI Frequency and latency.

Running MSR consistency test suite

(MSR 0x0 consistent): FAIL
(MSR 0x2a consistent): FAIL
(MSR 0x39 consistent): FAIL
(MSR 0x118 consistent): FAIL
(MSR 0x19e consistent): FAIL
(MSR 0x400 consistent): FAIL

Summary: 4687 passed, 6 failed


SMI Frequency and latency

Summary of impact: observed maximum latency = 647us
Risk of OS timeouts exists with this observed latency (ideal is < 150us)


This is with stock speeds and BIOS Defaults using BITS 316.
http://downloadcenter.intel.com/Detail_Desc.aspx?DwnldID=19763

It doesn't look that good does it?


Is this the reason why this board is so bad with Core2Quad processors?

Is this the reason why we've got these MACHINE_CHECK_EXCEPTION/0x...124
BSOD's?

Is this also the reason why I can't use Loadline Calibration without a
BSOD trying to boot into my OS? It seems to work fine with BIOS Defaults
but not when I need it to overclock my Q9650.


A lot of questions but you have to know that we are tired of these
issues with the board and are thinking to upgrade to X68 when it's
available. With all these problems that we've got with our ASUS board
it's very unlikely that this is going to be build on an ASUS board and ASUS
is going to lose hundreds if not thousands customers.


Why doesn't ASUS want to fix these issues that we're having with our
ASUS Rampage Formula? Why is it so hard to understand that we would like
an Official BIOS without these issues with updated Option ROM's and DRAM
Refresh Period, MCH Read ODT and maybe DRAM Read and Write ODT if this
would be necessary?


That the board is EOL is no excuse for ASUS so they don't have to do
anything about it.


Best Regards,
XXX

[greg.m]

I decided to give it an other try.

"That the board is EOL is no excuse for ASUS so they don't have to do
anything about it."

Besides, what if someone bought it last year for example? It still has 2 years warranty and by saying that, support is the least asus can offer!

[leo27]

I've also emailed ASUS with the results from the Intel BIOS Test Tool but it doesn't look very good. I get great response from the customer service contact, but it doesn't seem to go to the next level i.e. to the BIOS department.

[Alien Grey]

I've also emailed ASUS with the results from the Intel BIOS Test Tool but it doesn't look very good. I get great response from the customer service contact, but it doesn't seem to go to the next level i.e. to the BIOS department.

I think that the ASUS engineers are telling everyone at ASUS that there's nothing wrong with the BIOS or board because they don't want anyone to know that they screwed up or they don't want you to have anything stable running than default clocks.

But Intel did us a favor with releasing the BIOS Implementation Test Suite and now they can't deny it anymore and I really hope that someone very important at ASUS is now going to take the necessary steps to fix our BIOS issues with the board.

[JohnZS]

I agree Alien Grey

ASUS make great boards, but sadly the quality of their BIOSes since the days of the P4C800 are lacking.

ASUS Support at times do mean well, they managed to resolve my SSD RAID issue and have tried looking into the QX9650 VID Ratio error whenever I use a manual multiplier.
They gave me an internal BIOS which fixed the SSD's and had the latest revision Microcode for the QX9650 (60C as opposed to 60B) Sadly the VID Ratio issue STILL remains.

ASUS are out of ideas, I am out of ideas

I reported the BIOS Implementation Test Suite results, but ASUS seemed to ignore it.

Our latency issue could be significantly reduced by having the option to disable HPET.

John

[Alien Grey]

I agree Alien Grey

ASUS make great boards, but sadly the quality of their BIOSes since the days of the P4C800 are lacking.

ASUS Support at times do mean well, they managed to resolve my SSD RAID issue and have tried looking into the QX9650 VID Ratio error whenever I use a manual multiplier.
They gave me an internal BIOS which fixed the SSD's and had the latest revision Microcode for the QX9650 (60C as opposed to 60B) Sadly the VID Ratio issue STILL remains.

ASUS are out of ideas, I am out of ideas

I reported the BIOS Implementation Test Suite results, but ASUS seemed to ignore it.

Our latency issue could be significantly reduced by having the option to disable HPET.

John

I don't think that disabling HPET is going to reduce the SMI Frequency and Latency. It reduces the DPC Latency and that's not a problem on the ASUS Rampage Formula with the Supreme FX II soundcard.

Something very interesting is that the Intel DX38BT also fails the MSR Consistency Test. It also fails the MSR 0x0 and 0x39 so these ones aren't really an ASUS problem but look more like an Intel problem.


Anyway I'm not happy. I've spend a lot of money on hardware to build this computer and the fact is that it runs like and nobody at ASUS seems to care about this. Maybe they should have a good talk with Intel because I think that most of our problems are also their fault and to be honest, if it was me who has bought one of their Extreme Edition Quad processors and couldn't change the multiplier to boost the frequency, I would be pissed that I spend so much money a CPU that fails to do what it's supposed to do.

[greg.m]

Guys which are the best audio drivers to use?

[leo27]

@Greg M
The best audio drivers are the latest SoLoR modded ones v6610

Also, i got a reply from ASUS today. Bascially they aren't going to produce a new BIOS because they say that this board is old and on that basis isn't worth support.

They said they might implement our requested options on their new range, like anyone here would buy ASUS again.

[Counter CS]

@Greg M
The best audio drivers are the latest SoLoR modded ones v6610

Also, i got a reply from ASUS today. Bascially they aren't going to produce a new BIOS because they say that this board is old and on that basis isn't worth support.

They said they might implement our requested options on their new range, like anyone here would buy ASUS again.

ASUS sucks

[Alien Grey]

@Greg M
The best audio drivers are the latest SoLoR modded ones v6610

Also, i got a reply from ASUS today. Bascially they aren't going to produce a new BIOS because they say that this board is old and on that basis isn't worth support.

They said they might implement our requested options on their new range, like anyone here would buy ASUS again.

Maybe they can try to find a way to fit my Q9650 and my G.SKILL F2-9600CL5D-PI memory on their new range.

It's the most lame answer they could give you and to be honest I don't give a damn anymore if they release a new BIOS or not. It's obvious that don't want us to have a nice stable overclock with that board and buying one of their new range would be the stupidest thing you could. The support would still suck and before you know it you would end up with a Rampage Formula look a like.

[Counter CS]

But why do they continue updating bios frome the P5W DH Deluxe ?

It´s older than my maximus Formula

[Alien Grey]

I tested the MSR consistent with the first release BIOS 0219 and it also has the 6 fails so I think that every BIOS has these fails. No explanation so far from ASUS what they mean and if they're going to fix it.

[Counter CS]

I tested the MSR consistent with the first release BIOS 0219 and it also has the 6 fails so I think that every BIOS has these fails. No explanation so far from ASUS what they mean and if they're going to fix it.

Look at my post. I only have 4 fails.

MSR consistency test suite:

( MSR 0x39 consistent ) FAIL
( MSR 0x14e consistent ) FAIL
( MSR 0x1dd consistent ) FAIL
( MSR 0x415 consistent ) FAIL

Summary: 4851 passed, 4 failed...

Where can i get RF 1002 BIOS modded ?

[JohnZS]

Look at my post. I only have 4 fails.

MSR consistency test suite:

( MSR 0x39 consistent ) FAIL
( MSR 0x14e consistent ) FAIL
( MSR 0x1dd consistent ) FAIL
( MSR 0x415 consistent ) FAIL

Summary: 4851 passed, 4 failed...

Where can i get RF 1002 BIOS modded ?

I think it is because you are using a different CPU Counter CS.

I also noticed your post the other day and thought it was odd how me and Alien had 6 MSR fails on our respective boards with Quad Cores... so I did a test.

I ran the test suite using my QX9650 in my Rampage Extreme, and yep got 6 MSR fails.

Swapped the CPU for a E8400... guess what?

4 MSR Fails!

So ASUS seem to have some issues programming in the parameters for Quad Core CPU's as there are 2 more MSR fails when using a Quad Core compared to a Dual core.

I wonder if there would be even less MSR fails when using a 65nm "Conroe" Dual compared to a 45nm Penryn?

I would not be surprised if 65nm dudes have less MSR failures.. but we need to prove this.
John