ASUS Official X79 Motherboards Support Thread

[stasio]

82b,file incomplite.

[LardArse]

updated link
http://www.mediafire.com/?s5mbqsba3qjoa6p

[stasio]

Thanks.

[massman]

Anyone tested the Gskill F3-17000CL9Q-16GBXLD ?

- 4 sticks = stuck to B7
- 1 stick = boots fine
- 2/3 sticks = only shows 4GB in bootup/os

Dimm post tool in bios detects all 4 modules.

//edit: I'm on 0083b bios. Gskill RD tells me it should work.
//edit2: received new SPD, will test soon.

[pazza316]

My cold boot problem has gone at both 2133 and 2400mhz on the latest bios with all settings at AUTO and using 32gb memory. The memory is now running less aggressive timings as all benchmarks are down as is CPU temps. CPU temps down 3c for same voltage and GFLOPS down 4-5 on Linx, its similar for all other benchmarks.The AUTO settings set VCCSA at 1.35, VTT 1.35 for and PLL 2.0 for 2499mhz is PLL at this ok 24/7??

To try get some of the speed back I then used the hynix profile which is what my RAM is. On this profile I am unstable in wondows, with prime, linx etc without 1.71v on the RAM. I notice the profile is for 4x4gb I am using 8x 4gb. Raja or Shamino could you look at doing a 8 sticks profile or suggest to me the most prominent settings to tweak to regain stability but keep some of the speed this profile gives . With 4 sticks the profile works great and no cold boot. Thanks for the great support guys best around to be honest. Hey one more question the Bios above 82b Shimano posted what are the changes on this bios???

[LardArse]

My cold boot problem has gone at both 2133 and 2400mhz on the latest bios with all settings at AUTO and using 32gb memory. The memory is now running less aggressive timings as all benchmarks are down as is CPU temps. CPU temps down 3c for same voltage and GFLOPS down 4-5 on Linx, its similar for all other benchmarks.The AUTO settings set VCCSA at 1.35, VTT 1.35 for and PLL 2.0 for 2499mhz is PLL at this ok 24/7??

To try get some of the speed back I then used the hynix profile which is what my RAM is. On this profile I am unstable in wondows, with prime, linx etc without 1.71v on the RAM. I notice the profile is for 4x4gb I am using 8x 4gb. Raja or Shamino could you look at doing a 8 sticks profile or suggest to me the most prominent settings to tweak to regain stability but keep some of the speed this profile gives . With 4 sticks the profile works great and no cold boot. Thanks for the great support guys best around to be honest. Hey one more question the Bios above 82b Shimano posted what are the changes on this bios???

woah?
0083b + auto settings = cold boot issue
0901 + auto settings = no cold boot issue
0083b + manual settings = ??
0901 + manual settings = ??
that what you are sayin? i cant know what is OK, what is not OK without clearer details.

[pazza316]

Sorry Shamino this is what I have tried.

0083b + Auto = UNknown never tried it.
0901 + Auto = NO COLD BOOT and stable with 16gb and 32gb @ 2133 or 2400
0901+ Hynix Profile 16gb = NO COLD BOOT and STABLE at 2133 or 2400
0901 + Hynix Profile 32gb = NO COLD BOOT and Unstable without 1.71v RAM 2400. 2133 stable at 1.65V
0083b + Hynix Profile 16gb = Cold boot problem Randomly but once in Windows Stable.
0083b + Hynix Profile 32gb = Cold boot always Hynix profile stable but only at 2133mhz not 2400 mhz. 2400mhz requiring 1.725 RAM volts.

Thanks

[LardArse]

thanks i got it now
So conclusion is 0901 doesnt cold boot issue for you.
No matter what BIOS 8x4GB Hynix @ 2400 32GB is unstable BUT 2133 is stable
-> So this is no surprise. Still working on it but 8 sticks stable at 2400 I think ONLY GSkill 8GB DIMMs can miraculously do it.

BUT Why only Hynix profile tested? Dont your DIMMs come with XMP too?

[pazza316]

Yes they do come with XMP. I will try that tonight when I get in and see where I get. TBH I forgot about XMP in this latest round of testing. The Hynix profile offers more speed than the XMP so I was trying to get that one stable. Its strange that the 8gb dimms are stable at 2400 but teh 4gb not!! usually its the other way. Please post if you find another way to get 8 Dimms sable at 2400. I will try XMP as this may work.

[Raja@ASUS]

I don't think 8 of the DIMMs you have currently can be stabilized at DDR3-2400 without some serious performance hits. Anything over one stick per channel at DDR3-2400 needs so many timings relaxed that DDR3-2133 can be made to be faster on concurrent reads in many scenarios. Been working with some DIMMs over the past few days that are theoretically "better" than what you have there. Even with a LOT of the timings relaxed to max could not get them consistently stable at 2400 at CR2 let alone CR1. I know 2400 is a sexy lookin number, but it does not mean it's the fastest way to run your memory


I'll also post the information "template" again on this page, so you guys with cold boot problems can help us to help you:

Please list the following:

Motherboard:
CPU Model:
Memory model, include part number and number of modules:
PSU model:


BIOS Version:


CPU Frequency:
Memory frequency:
Clockgen Filter:
Vcore:
VCCSA:
VTT:
VDIMM:
CPU PLL Voltage:
Memory Timings
Rampage Tweak Mode
Any memory profile


Additionally, please list ALL settings you have changed from Auto.


Lastly, please also state the naure of the cold boot problem, ie freeze when loading Windows or failure to POST etc...

[pazza316]

OK cheers RAja. I accept what your saying. They run stable at 2400 with 4 dimms at tight timings so I will either sell 4 dimms or stick with 2133mhz with 8 dimms. You can get em stable at 2400 with arond 1.725v which is abit high for 24/7 use, but they are 100% stable.

[Raja@ASUS]

Okay cool. If I get time tomorrow I might see how tight we can take 8 of those sticks at 2133 and post up the settings. However, our first priority at this stage is to get solid cold boot info back from you guys so we stand a chance of replicating it.

[pazza316]

I will post you full info when I get home although the basics I have allready stated. Tight 2133, 8 Dimm timings would be great cheers.

[mdzcpa]

Raja, here is my template -


Please list the following:

Motherboard: Rampage IV Extreme
CPU Model: 3960X
Memory model, include part number and number of modules: G.SKILL Ripjaws Z Series 16GB (4 x 4GB) 2400 (PC3 19200) Model F3-19200CL11Q-16GBZHD
PSU model: Corsair AX1200


BIOS Version: 0604, 0703, 0803, 0901 (all exhibit the cold boot symptom)


CPU Frequency: 4.8Ghz
Memory frequency: 2133, 2400 both have issue, 1600 is okay
Clockgen Filter: Auto
Vcore: 1.402v (using Offset Mode +.080) ALSO used Manual Vcore up to 1.44v without effect.
VCCSA: Auto (1.3v manual reduces issue, does not eliminate)
VTT: Auto (1.3v manual reduces issues, does not eliminate)
VDIMM: 1.65v (per XMP)
CPU PLL Voltage: Auto
Memory Timings: Per XMP (11-11-11-28-128 2T)
Rampage Tweak Mode: Disabled
Any memory profile: XMP


Additionally, please list ALL settings you have changed from Auto.
AI Overclock Tuner: XMP and Manual (both exhibit issue, though XMP seems to reduce issue)
Turbo Ratio: All Cores
Ratio: 48
Memory Speed (down clocking XMP default of 2400 to 2133 or lower for stability)
Vcore Offset Mode + .080 AND Manual Vcore from 1.375v-1.44v
CPU fans 1 & 2 Ignore

Lastly, please also state the naure of the cold boot problem, ie freeze when loading Windows or failure to POST etc...
System fails to fully load to windows every 2 or 3 cold boots
Mostly a quick BSOD right after windows starts to load, occassionally a black screen fail to post, sometimes an Overclock Fail message.
However, once booted into Windows successfully, the above settings are completely stable and can withstand overnight LinX, Prime, and 3Dmark.

[LardArse]

woah i didnt know you were on offset mode for vcore.
That is so hard to grasp.
The cpu decides the base VID and we just add an offset to it, we have no control on what VID the CPU decides to spew out at different times and different loads. For sure i would use Manual fixed mode of 1.4+v first to eliminate this.

[CosmoJoe]

Hey folks. Hoping to get some advice on an issue I am having with brand new P9X79 WS build (BIOS 0603). Components are as follows:
CPU: 3930K
Cooler: Corsair H100
Video Card: EVGA GTX580
SSDs: Kingston HyperX SH100S3B x2, Crucial M4 CT064M4SSD2 (Boot SSD)
RAM: Ripjaw Z DDR3 1600 Memory Model F3-12800CL9Q-16GBZL (x2 16GB kits so all banks are filled)
PSU: Corsair HX1050

Symptom: Random lockups in Windows and in event viewer, noticing frequent crashes of display driver ("nvlddmkm stopped responding and has successfully recovered"). For example, at one point I just changed the theme in Windows; the moment the wallpaper changed over, the screen went black with some random pixels where the start button should have been. After leaving the system run overnight I could not get the monitors to come out of power save mode and had to force reboot; noticed in event log multiple occasions of driver crashing.

I am going to swap out my video card this afternoon with a spare I have, but I had no such issues with this card in the system I am upgrading from so I am not sure how it could have failed just from switching it over, unless the card was on its last legs or something.

Any other suggestions as to what might cause these symptoms? I have run memtest86+ and it completed in about 3 hours with no errors which seems to indicate I cross the RAM off the suspect list.

[mdzcpa]

woah i didnt know you were on offset mode for vcore.
That is so hard to grasp.
The cpu decides the base VID and we just add an offset to it, we have no control on what VID the CPU decides to spew out at different times and different loads. For sure i would use Manual fixed mode of 1.4+v first to eliminate this.

Sorry, I should have been more clear in my post. I have also tried fixed Vcore up to 1.44v without effect. I listed my Offset settings as that was my preferred option for better thermals and cpu longevity. I have amended my post for this new info.

[LardArse]

got it , ive got some other guy testing out the test bios so maybe i will post 0085b if he finds it helps

cosmo realy sounds like the VGA card, maybe reseat cpu and vga in another slot (both could be suspects since lanes run from cpu to card)

[CosmoJoe]

got it , ive got some other guy testing out the test bios so maybe i will post 0085b if he finds it helps

cosmo realy sounds like the VGA card, maybe reseat cpu and vga in another slot (both could be suspects since lanes run from cpu to card)

Thanks, yeah I am going to try that out.

On a related note, does anyone here use SpeedFan? I downloaded the latest version and it causes a Windows crash and reboot for me.

[pazza316]

I have to add to my previous post that if I use XMP on 0901 or 0083b with 8 sticks I get my usual cold boot problem. My cold boot problem in all cases is Windows stops loading at the Splash Screen just freezes. So Shimano yes I tried XMP but its no good in my case.

However what is good is setting Clock gen to 20uf and XMP then no cold boot with 8 sticks. I tried this as you once suggested it for a previous problem and it worked well.

[Cookiesowns]

I have to add to my previous post that if I use XMP on 0901 or 0083b with 8 sticks I get my usual cold boot problem. My cold boot problem in all cases is Windows stops loading at the Splash Screen just freezes. So Shimano yes I tried XMP but its no good in my case.

However what is good is setting Clock gen to 20uf and XMP then no cold boot with 8 sticks. I tried this as you once suggested it for a previous problem and it worked well.

I've gotten the cold boot issue before, first time was related to a unstable system / loose power cable. Re-mounted everything and bam fixed, second cold boot was a RAID SSD issue, updated firmware, re did array, perfect. But this was on X58, so take it with a grain of salt. Not sure if it's relevant or not, but just trying to help!

[Raja@ASUS]

Raja, here is my template -


Please list the following:

Motherboard: Rampage IV Extreme
CPU Model: 3960X
Memory model, include part number and number of modules: G.SKILL Ripjaws Z Series 16GB (4 x 4GB) 2400 (PC3 19200) Model F3-19200CL11Q-16GBZHD
PSU model: Corsair AX1200


BIOS Version: 0604, 0703, 0803, 0901 (all exhibit the cold boot symptom)


CPU Frequency: 4.8Ghz
Memory frequency: 2133, 2400 both have issue, 1600 is okay
Clockgen Filter: Auto
Vcore: 1.402v (using Offset Mode +.080) ALSO used Manual Vcore up to 1.44v without effect.
VCCSA: Auto (1.3v manual reduces issue, does not eliminate)
VTT: Auto (1.3v manual reduces issues, does not eliminate)
VDIMM: 1.65v (per XMP)
CPU PLL Voltage: Auto
Memory Timings: Per XMP (11-11-11-28-128 2T)
Rampage Tweak Mode: Disabled
Any memory profile: XMP


Additionally, please list ALL settings you have changed from Auto.
AI Overclock Tuner: XMP and Manual (both exhibit issue, though XMP seems to reduce issue)
Turbo Ratio: All Cores
Ratio: 48
Memory Speed (down clocking XMP default of 2400 to 2133 or lower for stability)
Vcore Offset Mode + .080 AND Manual Vcore from 1.375v-1.44v
CPU fans 1 & 2 Ignore

Lastly, please also state the naure of the cold boot problem, ie freeze when loading Windows or failure to POST etc...
System fails to fully load to windows every 2 or 3 cold boots
Mostly a quick BSOD right after windows starts to load, occassionally a black screen fail to post, sometimes an Overclock Fail message.
However, once booted into Windows successfully, the above settings are completely stable and can withstand overnight LinX, Prime, and 3Dmark.

Thanks for the info. Does this happen at lower CPU frequenices also (try 43~47X)?

-Raja

[mdzcpa]

Thanks for the info. Does this happen at lower CPU frequenices also (try 43~47X)?

-Raja

45 and above appear to be problematic. 42-43 seem fine...though the issue is very random and I'm not sure I cycled enough times at lower frequencies to fully conclude. The longer between cold boots the more prevalent the issue, so it takes a long time to cycle. Most common occurence is when the system remains off overnight or all day.

However, good news, I applied some of the adjustments noted above a few hours ago, and I think the cold boot issue may be resolved for my configuration.

I applied the following settings for a CPU OC at 4.8Ghz with the ram at 2133mhz on UEFI 0803:
AI OC Tuner: Manual (changed from XMP or Auto)
Turbo Ratio: All Cores
All Cores: 48
Clock Gen Filter: 20UF (changed from Auto)
Memory Fequency: Manually set to 2133
Vcore: Manually to 1.42v (changed from offset mode or Auto)
VTT: Manually set to 1.25v (changed from Auto)
VCCSA: Manually set to 1.285v (changed from Auto)
RAM: Manaully set to 1.675v (changed from XMP of 1.65v)
PLL: Manually set to 1.95v (changed from Auto)
All other voltages and settings on Auto
Set DRAM timings manually to 11-11-11-31-2T (same as XMP but entered manually...all other settings on Auto).

I had previously set both vccsa and vtt to 1.3v (with all other settings on XMP and Auto except Vcore) which cut down on the occurence but did not eliminate the issue. The manual settings above appear more stable at boot and it's preferred with lower vtt and vccsa needed.

I have cycled about 12 cold boots now with 5-30 minutes in between and so far so good.

[christefan]

To Raja or LardArse---Got my replacement Rampage and my replacement CPU, my first board and processor choked when I tried to run 4 cards in the pci e slots. I have the system reassembled and are working on installations and tuning. Current problem is if I connect a SSD (OCZ Agility3) to the eSATA port it won't quick format the drive, that operation fails. The other big issue is I can't get SLI to work. Configuration is GTX580, GTX580, GTX560TI, and Areca 1680ix Raid card. I have a single sli connector from one 580 to the other and them another single sli connector from the 560ti to the second position on the first 580--the Nvidia control panel wants to 'auto select the 560ti for physics and enable sli on the two 580s-when I click apply it doesn't take and I'm left with sli disabled and physics jumps to cpu for physics--I don't get what's going on. BTW if I boot with something connected to the eSATA ports then I lose the bios screen for the Areca card---any help with these issues will be appreciated. I have to say you guys are doing a great job and I tried a Gigabyte board to see what was up with their x79 units--terrible, not even usable at this point with how bad their bios is--thanks

[LardArse]

christefan
you should not be hooking up the sli connector from 580 to 560.