Overclock.net › Forums › Intel › Intel Motherboards › Asus Z9PE-D8 Owner's thread
New Posts  All Forums:Forum Nav:

Asus Z9PE-D8 Owner's thread - Page 4

post #31 of 2450
anubis1127,
I had those symptoms, plus issues with my PCIe slots. In the "first week of life" I found that swapping RAM sticks in their slots did provide some relief of the posting inconsistencies; but I also had symptoms of my monitor not getting a DVI signal from my GPU. Turns out my CPU coolers were a bit too tight, causing unstable seating of CPU(s) in their seats. Symptoms of that were intermittent problems with posting and my graphics card output. Once that CPU cooler issue was solved (second week of life), I began to be able to post consistently and boot fairly consistently. I wanted so bad to believe that my troubles were over...
If I was to do it all over again, I would definitely employ MemTest to check the RAM. I'd do it by building a boot-able MemTest USB drive, then boot the system into it and test the RAM with only 1 stick installed at a time. That approach would have prevented a lot of headaches for me... If you want to read a summary of those headaches, so you have something to laugh about, read further... or not...


In the "second week of life" I decided to try to finish installing little things like my new ASUS wireless PCIe card. Unfortunately that caused Windows to freeze or BSOD, no matter which slot I chose. ASUS RMA'd it but the replacement had the same results.

In week 3 I started to do a lot of benchmarking/monitoring to get a feel for what kind of machine I had created. I employed free apps like PCMark, 3DMark, Prime95, NovaBench, etc. My system seemed to be just as beastly on 3D and rendering as I was expecting it to be. I did some burn-ins with some of the apps; but nothing formal or extensive. But I was beginning to discover errors and corruption in some of my 3dsMax and After Effects files. And Photoshop got corrupted (not normal). And Speccy showed some wierdness with the very same RAM sticks that I had swapped weeks earlier - showing only half of the RAM in those slots while the BIOS was detecting all of it! That was weird and disturbing. So now knowing that the CPU seating wasn't an issue, I tried re-swapping RAM to the original setup. That fixed the Speccy detection issue and luckily there were no boot issues. But it was clear something was up with the RAM and that usage was possibly magnifying those issues. As I did more research I discovered another recommended app called "SiSoftware Sandra". I installed that; but it couldn't even get through a memory test without freezing up. Hmmm...

In week 4, I began to experience more and more problems with booting. Occasionally it would freeze during the initial Windows splash screen. And when that happened it damaged Windows. A few times I was prompted to have it try and repair itself, and a few times that was actually successful. But I began to need to rely on BU Images to recover. It was clear that something was not only wrong but seemingly getting "wronger". It made sense that my BU images were probably at least marginally corrupted by bad RAM. I got an email message that 3DMark11 had been updated and I tried it. It all seemed to go fine; but when it took me to the Futuremark site to post my results, there was an error message indicating that some corruption had prevented my results from being posted. A FutureMark tech told me that the accompanying error code would be indicative of a hard drive or RAM problem.
I was at a point where I began to question all over again which new parts in my installation were actually problem free. But I was also strongly suspecting that bad RAM could logically account for all of the continuing and growing issues.
So in week 5 I began to experiment with MemTest. I installed it and gave it a test drive by initiating a test on all of the installed RAM (64GB) at once. While a shotgun approach such as this is not the recommended approach for any meaningful results, I did quickly begin to see some red-lettered verification that there were some problems. It was obvious that the recommendation leaned more toward building a MemTest boot disk and doing a methodical series of tests on one stick at a time; but I decided that before I went to so much trouble, I would first see what I could see by stressing it out with that full RAM test and doing it overnight. Now the next morning I was a little preoccupied and forgot that I had turned off my monitors for the test. So I mistakenly thought my machine had frozen again. I did a hard shutdown and reboot, Windows was able to repair itself, and while searching the Windows logs for any record of the MemTest results, Windows froze. After that whenever I attempted to restart, it would only give me the BSOD finger. I removed the 32GB RAM kit with the suspect sticks, and PHEW!... it booted right up again. I concluded that the overnight test had stressed the marginal RAM to the point of imminent failure.

Later that week Windows got corrupted again. Rather than going back to my flakey BU image, I sought MS tech support for a repair solution that I had seen advertised. A few hours later it was clear that they could do nothing for me. So I decided to bite the bullet, wipe the drive and reinstall everything. But first, I made sure that I created a boot-able MemTest USB drive to do some meaningful testing. I booted from that and tested all of the RAM by installing/testing only one stick at a time - the way I should have from the beginning! Results showed that the kit I had removed earlier was suffering from 2 extremely bad sticks. And it showed that the 2nd kit had a marginally bad stick. So I reinstalled two good sticks, reinstalled all my software, and contacted Corsair about RAM replacements. Since then all has been stable. Corsair was kind enough to replace both of my 32GB kits and replace them with a single 64GB kit. I was grateful for that!

So 6 weeks from the "start of life", there was no longer any trend towards corruption, posting/booting is stable and dependable, and the consistent beastliness of my dual CPU monster is getting boring... Just how it should be. Maybe it's silly superstition; but I have not tried installing my ASUS wireless card or re-installing Sandy...
post #32 of 2450
I tried loosening the CPU cooler to the point where its barely tightened, but no change. I'm down to testing with just one CPU, and one dimm slot filled, and still no luck. I've tried unregistered dimms that are on the QVL, and two sticks of Samsung that isn't on the list, and neither made a difference.

I'm starting to wonder if my ES chips aren't compatible with the board, or if there is just an issue with the board. Unfortunately I don't have any other 2011 Xeons to test with the board. The CPUs work fine in my x79 desktop motherboard.
post #33 of 2450
Anubis1127,

You may have something there about the ES chips... there's not much guarantee that they wouldn't have a minor bug or two baked in. And as touchy as the Z9PE-D8 board seems to be, I could easily picture it choking while the less complex x79 works fine with it.

With the steps you took with the RAM and CPU it sounds like either your MoBo has issues or your ES chips are outside of its spec.

It's more than remotely possible that there is a little variation between your ES chips. In that case, you've probably thought about when you're doing the single CPU test, that you're giving each one a turn at it.
ASUS support seemed quite willing (eager even) to RMA my board... if that step interests you at all... It only interested me as a last resort and fortunately I didn't have to take that step.
Now I get your earlier comment about your RAM potentially out-pricing the CPUs...

One other thing... I hate to admit that I did this; but the very first time I tried to post, the Z9 MoBo had insufficient power from the PSU to do so successfully.
It took a little sleuthing before I realized that this board required the normal 24 pin ATX cable plus TWO of the 8-pin auxiliary power cables. I had connected only one of the 8-pins. Not saying you'd do that; but just goin' down the list of easy stuff to check.
Edited by BlackenedTush - 11/15/12 at 8:02pm
post #34 of 2450
Quote:
Originally Posted by BlackenedTush View Post

Anubis1127,

One other thing... I hate to admit that I did this; but the very first time I tried to post, the Z9 MoBo had insufficient power from the PSU to do so successfully.
It took a little sleuthing before I realized that this board required the normal 24 pin ATX cable plus TWO of the 8-pin auxiliary power cables. I had connected only one of the 8-pins. Not saying you'd do that; but just goin' down the list of easy stuff to check.

Bolded for reference, my board would not even attempt to power on without both of the 8-pin EPS plugs. Also, it did not like the 6-pin PCI-E to 8-pin converter at all, and I ended up buying an NZXT hale90 with 2 EPS plugs on it and it was good to go after that.
post #35 of 2450
I just picked one of these up "open box" for $375. thumb.gif we'll see how "opened" it was first before I purchase anything for it rolleyes.gif. I've been eyeing this board up for a very long time and way back in 2010 I started building up for a dual socket, so I'm a bit excited here smile.gif I really don't anticipate any problems, I've always had great luck with ASUS boards. This will be dropping into my main rig which I use for a VM host for a highly threaded software development project. I've had this i7 950 pegged while running tests for the past couple days and the RAM maxed, so, time to upgrade.

I'm looking at the E5-2687W as the CPU option, starting with just one for a period of time, see how it goes, and then a second for a kick in the nuts. I would think two of those suckers would laugh at this workload. If only the multiplier was unlocked .. come on intel redface.gif

I have the manual here to read and based on what I've skimmed through here, I have some careful shopping for RAM to do. I could see this touching 32GB, so I'll probably be in the 48 or 64 range to start. I don't understand, just a couple of years ago, 8GB DIMMs were always super super expensive, and now they are dirt cheap. ?

It had better be fully VT-D compatible this time too, the marvell chipset scares me as the one on my current board is out of spec, but I believe it can be turned off on this one. Is anyone here using the VT-D to passthrough a PCI card to a VM?
Edited by lloyd mcclendon - 11/23/12 at 12:11am
stable again
(25 items)
 
  
CPUCPUMotherboardGraphics
E5-2687W E5-2687W ASUS Z9PED8-WS EVGA GTX 570 (Linux host) 
GraphicsRAMHard DriveHard Drive
EVGA GTX 970 FTW (win7 guest) 64GB G.SKILL 2133 2x Crucial M4 256GB raid1 4x 3TB raid 10 
CoolingCoolingCoolingCooling
2x Apogee HD  2x RX 480 2x MCP 655 RP-452x2 rev2 (new) 
CoolingCoolingOSOS
16x Cougar Turbine CFT12SB4 (new) EK FC 580 Gentoo (host) Gentoo (x23 guests) 
OSMonitorMonitorPower
windows 7 (guest w/ vfio-pci) Viewsonic 23" 1080P Viewsonic 19" Antec HCP Platinum 1000 (new) 
CaseOtherOther
Case Labs TH10 (still the best ever) 2x Lamptron FC-5 IOGEAR 2 way DVI KVM Switch 
  hide details  
Reply
stable again
(25 items)
 
  
CPUCPUMotherboardGraphics
E5-2687W E5-2687W ASUS Z9PED8-WS EVGA GTX 570 (Linux host) 
GraphicsRAMHard DriveHard Drive
EVGA GTX 970 FTW (win7 guest) 64GB G.SKILL 2133 2x Crucial M4 256GB raid1 4x 3TB raid 10 
CoolingCoolingCoolingCooling
2x Apogee HD  2x RX 480 2x MCP 655 RP-452x2 rev2 (new) 
CoolingCoolingOSOS
16x Cougar Turbine CFT12SB4 (new) EK FC 580 Gentoo (host) Gentoo (x23 guests) 
OSMonitorMonitorPower
windows 7 (guest w/ vfio-pci) Viewsonic 23" 1080P Viewsonic 19" Antec HCP Platinum 1000 (new) 
CaseOtherOther
Case Labs TH10 (still the best ever) 2x Lamptron FC-5 IOGEAR 2 way DVI KVM Switch 
  hide details  
Reply
post #36 of 2450
Just got a Z9PE-D8 WS today. My issue is similar in terms of no post (not even fan spin or post-code), but there is a difference on mine. I have a new Seasonic 1000w Platinum + Z9 + 2x E5-2690. I've tried 1 million different combinations of things, but have stripped it down to 1 cpu (heatsink + fan) + power supply. You'd think the fans would at least spin or maybe a give me an error code, but nothing except for a really strange light pattern. The on-board power button turns red, then goes off, the onboard reset button turns green then goes off, then a small green led in the bottom corner goes on then off. The cycle repeats over and over. This is when power is simply attached to the board (before I even try to turn it on). This is sitting on a box on my bench with literally nothing else attached. I cant figure it out for the life of me. I'm thinking maybe a fried bios (since if i remove the bios clr jumper the pattern stops). I've taken away power, pulled the cmos bat, jumpered the clr pin then put it all back and the exact same thing. Completely frustrating and making me hate ASUS more and more by the second. Anyone else seen this or have ideas. I plan to RMA with Asus on Monday, but was hoping someone had an idea before it lose another week to shipping.
Edited by eq2account - 11/23/12 at 2:19pm
post #37 of 2450
lloyd,

I didn't anticipate any problems either; but I've had my share rolleyes.gif. This thread is full of em. But I'm happy with my board now that things are stable.

Even if unlocked, that E5-2687W is a pretty dang beefy processor (and quite a pricey "kick in the nuts" at that).

I had some issues with the Marvell chipset (can't quite recall right now what exactly it was). I've got 9 disks (correction - 10). I elected to keep my VT-D disabled. And come to think of it, I think I might have left Marvell enabled; but I'm actually avoiding it by employing a PCIe RAID controller for 5 HDDs and then using Intel's SATAs to cover 3 SSDs & 2 HDDs.

Sounds like a good choice to go with a 48GB or 64 GB kit to start with. Can't easily mix and match DIMMS these days, so adding RAM later can be problematic (unless of course you are willing to up the voltage and drop the frequency).

Hey good luck!

BlackenedTush
Edited by BlackenedTush - 11/23/12 at 7:35pm
post #38 of 2450
eq2account,

Sorry about your no post... and thanks for posting... (sorry, I couldn't resist).

I'm having a little difficulty understanding the sequence of what you've tried and what exactly happens as a result; but if the lights are going crazy before ever hitting the power button, it sounds to me like you have a goofy MoBo. But I also wonder if the clr jumper might be the culprit... maybe having come with the clrtc jumper installed in the wrong position to start with, and then after attempting the purge of the CMOS, you have put it back where it started.

Also, you didn't mention RAM when describing your stripped down config. Are we to understand that your stripped config doesn't include it? If so, I can't imagine it posting without it.

And also, you do you have three AIPX power connections from the PSU (two 8-pin and one 24-pin), correct?

BlackenedTush
post #39 of 2450
well, I got the board today, this thing is pretty heavy and big. very nice. looks great in person. The manual is tvery hick and there's a boatload more features to understand with it than the boards I am used to.

unfortunatley due to it being open box I am missing the rear IO shield. ..... any ideas for obtaining one? perhaps it was a member here that sent it back and has it somewhere. smile.gif

I picked up this ram which should work well: G.SKILL Ripjaws Z Series 32GB (4 x 8GB) DDR3 2133 (PC3 17000): http://www.newegg.com/Product/Product.aspx?Item=N82E16820231524

later i'll pick up a second kit along with the second CPU. I have to pick one of them up first ... but I've yet to convince myself that spending that much on a CPU is a good idea. rolleyes.gif But no turning back at this point redface.gif
Edited by lloyd mcclendon - 11/30/12 at 4:13pm
stable again
(25 items)
 
  
CPUCPUMotherboardGraphics
E5-2687W E5-2687W ASUS Z9PED8-WS EVGA GTX 570 (Linux host) 
GraphicsRAMHard DriveHard Drive
EVGA GTX 970 FTW (win7 guest) 64GB G.SKILL 2133 2x Crucial M4 256GB raid1 4x 3TB raid 10 
CoolingCoolingCoolingCooling
2x Apogee HD  2x RX 480 2x MCP 655 RP-452x2 rev2 (new) 
CoolingCoolingOSOS
16x Cougar Turbine CFT12SB4 (new) EK FC 580 Gentoo (host) Gentoo (x23 guests) 
OSMonitorMonitorPower
windows 7 (guest w/ vfio-pci) Viewsonic 23" 1080P Viewsonic 19" Antec HCP Platinum 1000 (new) 
CaseOtherOther
Case Labs TH10 (still the best ever) 2x Lamptron FC-5 IOGEAR 2 way DVI KVM Switch 
  hide details  
Reply
stable again
(25 items)
 
  
CPUCPUMotherboardGraphics
E5-2687W E5-2687W ASUS Z9PED8-WS EVGA GTX 570 (Linux host) 
GraphicsRAMHard DriveHard Drive
EVGA GTX 970 FTW (win7 guest) 64GB G.SKILL 2133 2x Crucial M4 256GB raid1 4x 3TB raid 10 
CoolingCoolingCoolingCooling
2x Apogee HD  2x RX 480 2x MCP 655 RP-452x2 rev2 (new) 
CoolingCoolingOSOS
16x Cougar Turbine CFT12SB4 (new) EK FC 580 Gentoo (host) Gentoo (x23 guests) 
OSMonitorMonitorPower
windows 7 (guest w/ vfio-pci) Viewsonic 23" 1080P Viewsonic 19" Antec HCP Platinum 1000 (new) 
CaseOtherOther
Case Labs TH10 (still the best ever) 2x Lamptron FC-5 IOGEAR 2 way DVI KVM Switch 
  hide details  
Reply
post #40 of 2450
smile.gif You couldn't help posting. Good one.


I actually RMA'ed the board back to NewEgg as Asus was taking way way too long for their "Advanced Replacement" process. New Egg was out of stock on Z9 so I just bailed on them altogether. I found a used Z9 (that I had the owner send me pictures of actually running) and my problem was solved. The only thing I am wondering about now is the 0503 bios. There are newer ones, but 0503 is working and I can see that just a slight mistake or reset in bios means clearing bios and starting over. For me that means taking the case apart and I'm not in the mood. Is anyone running the 1107 bios and can anyone think of a good reason to upgrade? I'm not convinced turbo mode is working flawlessly, but i do see that the cpu jumps up to 3.1 (from 3.0) here and there. I have E5-2687w and I was hoping it would hit 3.6 or 3.5 GHz once in awhile. Is there a mb monitor tool like there are for the ASUS RoG boards that will let me check temps and fan speeds and clock?

The original issue with the RMA'ed board was determined to be bios corruption by Asus Tech Support. I got $575 back and picked up the working (but used) board for $300, so it felt like a slight win.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel Motherboards
Overclock.net › Forums › Intel › Intel Motherboards › Asus Z9PE-D8 Owner's thread