I appears that I'm having difficulty establishing even a mid-level steady overclock.

 

What were your BIOS settings?

 

Did you get your clockspeed stable first before touching uncore etc?

Haven't messed with uncore yet ... had a really steady OC at 1.29Vcore with a 44x multiplier but it wasn't as rock solid as I thought. 

 

It would appear that I'm good for about five to six hours then Prime95 says "No More" ... AIDA and OCCT were good ... but not forever.

 

The motherboard is beginning to exhibit signs of problems ... when stress testing it will not keep time ... and the Ethernet adapter (on board) continuously must be reset.  (A little yellow yield icon appears in the status bar, right clicking on it and telling Windows to fix it always results with a good connection with a new IP ... problem is that I control my mouse and keyboard on another computer via a program called Synergy ... every time the IP address changes the relationship between server and host must be redone ... PITA).

 

Going to contact tech support today when they open ... they might tell me to return it to Newegg ... even more of a pain in the A$$ ... will keep thread updated. 

 

Thanks for all the help guys ... this has been one he11 of a learning experience.

Take for example at 4.6Ghz I would pass everything else but Prime95 would fail after 4-5 hours.

 

My experiance is that Prime need 0.02-0.05V more Vcore than Aida FPU and other tests to pass,  passed prime at 4.8 for 12hour now

the AIDA FPU temps raised 3-5 c on the edge with fpu test but hope it be good with CLP think a can forget 100% stable @4.9 even with CLP.

But FSX runs @ 5.0ghz stable, 4.9 in FSX would be without any problem as safe fsx clock at same vcore as @4.8 100% Aida fpu , prime95 . OCCT.

Have not test the hardest AVX2 test that Jimmy G have a link to .

Well, after time on the phone with ASUS technical support ... it would appear that the problem my motherboard is having with keeping time is a BIOS chip problem.  So, off it goes to ASUS ... 2-1/2 weeks later ... my new board will return and guess what ... I get to start all over again!  I was TWO days over the 15 day return policy that Micro Center has regarding motherboards ... I'm such a lucky guy.

 

I'm so ecstatic!

 

Also, when I realized what was happening and ASUS confirmed it, I asked them about a problem I was having with an unrecognized device, Microsoft ACPI system of some sort.  They said that particular driver has to do with the power plan options in Windows ... which would explain why, everytime I booted back into windows, my Power Plan settings of Never and Never always reverted to 1 hour and 2 hours.

 

The technician said he fixed this one time ... but couldn't remember how he did it ... and that the best way to do it was to complete a re-install of Windows.

 

I then asked whether or not installing the Chipset driver from the files available on the website first instead of using the CD that came with the motherboard could have caused a problem.  He assured me that no, that would not be a problem. 

 

Soo ... I re-installed Windows (why not, my motherboard is gone anyway), reinstalled the Intel Chipset from there and low and behold, what disappeared from the device manager?  Yep, the Microsoft ACPI somethingorother was no longer an unrecognized device.  So, it would appear that NickN is absolutely correct in the following regard:

 

1.  Always install the drivers from the CD that came with the motherboard first.

2.  Install the Intel-Chipset_Win7_8 Drivers FIRST

3.  Install the Intel_MEI_Win7_8 SECOND

4.  Next, download the those updated drivers from the website

5.  Install the updated Chipset drivers first

6.  Install the updated MEI drivers next

7.  Continue on with other drivers as you see fit

 

REBOOT after each dirver installation ...

 

See you guys in 2-1/2 weeks ...

Another app for testing the stability of OCs is Intel Burn Test. It's  a GUI for the Intel LINPACK libraries and it's what I have used:

 

http://www.intelburntest.com-about.com/

 

It's a bit harder to configure, but it does an excellent job of stressing all your components, which is really what are are shooting for with a stable OC. The most recent version is 2.54 (27 March 2013).

...

 

So, it would appear that NickN is absolutely correct in the following regard:

 

1.  Always install the drivers from the CD that came with the motherboard first.

2.  Install the Intel-Chipset_Win7_8 Drivers FIRST

3.  Install the Intel_MEI_Win7_8 SECOND

4.  Next, download the those updated drivers from the website

5.  Install the updated Chipset drivers first

6.  Install the updated MEI drivers next

7.  Continue on with other drivers as you see fit

 

REBOOT after each dirver installation ...

Oh yes! The procedure after a fresh Windows install is very, very important. I wish that the ASUS DVD that is shipped with the mobo, would have a wizzard to ensure that a user installs the proper drivers in the correct order.

 

 

I'm not sure if all owners of the Samsung 840 SSD is aware of this info that Nick has updated his "bible" with:

 

IMPORTANT INFORMATION: Samsung SSD drives

 

Samsung has defined that if you are using their SSD drives that the Intel RST (Rapid Storage Transport) driver should NOT be installed.

 

PLEASE be aware of this and also be aware that if you are mixing Intel and other SSDs that may need RST driver support with Samsung drives you may want to reconsider that move so your Samsung SSD follows the manufacture recommend path with AHCI SATA drivers.

JaBloom:

 

I used Burn Test ... selected 15 passes as well as the Very High option and the right clicked on Start and selected Extreme.  I passed those 15 tests easily ... and failed Prime95 after 5 to 7 hours.

 

I guess at some point it is time to say enough is enough and accept something in the middle instead of going to a Prime95 of 24 hours. 

 

Maybe it was my motherboard ... with a defective BIOS chip it may have caused some stability issues ... who knows.

 

At least now I know what to do and why ... it gave me the opportunity to try Liquid Electrical Tape on the SMB components to the left of the i7 Thermal plate as well as get rid of the MX-4 ... which was behaving admirably by the way ... in favor of CLP from top to bottom.

 

Watch my CPU fail after I get all of this back together again ... sheeesh!

 

Ulf:

 

Dude ... THAT was what I was trying to remember and the topic of a new post I placed on the AVSIM forum for Nick ... I knew there was a driver that shouldn't be installed with those SSDs ... I just couldn't remember which one.  Now I wonder if my two WD Caviars meet the criteria ... I'll research that and find out.

 

CT

CT,

 

Hope that you get the problem with the mobo sorted out! DMZ just posted an interesting answer to your post over at SimForums.

 

 

IMPORTANT INFORMATION: Samsung SSD drives

Samsung has defined that if you are using their SSD drives that the Intel RST (Rapid Storage Transport) driver should NOT be installed.

PLEASE be aware of this and also be aware that if you are mixing Intel and other SSDs that may need RST driver support with Samsung drives you may want to reconsider that move so your Samsung SSD follows the manufacture recommend path with AHCI SATA drivers.

 

I'm using RST w/ a Samsung 840 and don't see any issues so far.  How would I know?  I think I remember the Win 7 Performance rating was as high as it can get for storage.

I'm using RST w/ a Samsung 840 and don't see any issues so far. How would I know? I think I remember the Win 7 Performance rating was as high as it can get for storage.

I use Intel SSDs and have no idea what the problem might be. I just noticed Nicks warning about using Intel RST when using Samsung 840. I would contact Samsung support and get the info from their tech experts.

I used Burn Test ... selected 15 passes as well as the Very High option and the right clicked on Start and selected Extreme.  I passed those 15 tests easily ... and failed Prime95 after 5 to 7 hours.

 

I know this is against the conventional wisdom, but I have a different take on the whole issue of evaluating system stability by running the multicore routines that arguably stress your system more than FSX does.  I know heat generation is any indication, temp definitely goes up significantly w/ Prime95 over what it does w/ FSX in my build.  JJ at Asus in his video re overclocking SB-E mentions this issue and I think it makes pretty good sense.  Why not set up the machine to run FSX as the burn in routine?  If the logic holds true you should end up w/ lower voltages & temps to get FSX to run stable compared to what might be required for these synthetic benchmarks to maintain stability.

That is certainly a novel idea. The first issue with that approach, is most people do use their computers for running apps other than FSX. The idea makes sense if your mobo has multiple BIOS profiles, as you could just boot up into the "FSX" OC profile and you probably would be okay.

 

The second issue would be how exactly to use FSX as the burn-in app. There are so many possible combinations of aircraft, scenery, weather and AI Traffic, that one might get a stable result one day and then not the next. It would probably take a lot more tuning than using Prime95, AIDA or IntelBurnIn.

 

I suppose as an alternative that you also could limit your burn-in app to the same number of physical cores that FSX supposedly uses by explicitly setting the app's affinity.

 

 

That is certainly a novel idea.

 

I'm essentially doing a hybrid approach:  I run Prime95 for 10 minutes, lowering voltages (VCORE & VCCSA) incrementally until I get an error.  Up one notch and jump into FSX and try a long flight and see how she goes.

Ah man ... I just had everything installed too.  Now I'm going to have to go through the process of re-installing FSX Acceleration (via the phone method) ... ask PMDG to give me another activation ... hours and hours of re-reading Nick's instructions on setting up Windows ... about five programs from Flight1 ... sooner or later they'll start to ask what the heck is going on ... sigh.

 

As for using FSX to burn in ... I cannot attest to that ... especially since on those long flights of mine I'm at FL360 and the frames are way up there anyway ... but when I first started this test and Noel asked me to max everything out and go to KSEA, 34R, Major Thunderstorm ... I did get a blue screen ... and had to up Vcore and go around again.

 

Also, I just read through the warranty agreement on ASUS website and it stated that they may repair or replace ... their discretion ... I have to send the motherboard in with original packaging ... if not they may charge me for packaging upon return ... and that the original packaging that I have to send the product to them in may or may not be returned to me.  Man, why can't they just yank another one off the shelf and be done with it?  I mean I paid $219 for a motherboard ... why split hairs trying to save money when they already wasted 2-1/2 weeks of my time (3-1/2 weeks if you count trying to figure out what was wrong with the thing)!

 

I'm pretty ###### now ... maybe they'll surprise me ... doubt it though.

 

 

I'm pretty ###### now ... maybe they'll surprise me ... doubt it though.

 

Best of Luck CT that's a bummer!