How to Overclock Q9550 on EP45T-UD3R with DDR3-1600

[profJim]

Re: How to Overclock Q9550 on EP45T-UD3R with DDR3-1600

The AUTO setting for the primary memory timings is probably the cause for the cas latency = 10 setting.

The BSOD might have been due to running ET6 while you were running Prime95.
Years ago some forum members reported strange system behavior while ET6 was running.
I never let ET6 auto start on next reboot and I only have ET6 running while I create a screenshot.

Check to see if you are using the latest version of ET6 by right clicking on the ET6 icon and selecting Open file location.
Open the verinfo.ini file by double clicking on the file and Date=2012/11/21,0 will be near the beginning of the file if you have the latest version.
Go to GIGABYTE - Motherboard - Socket 775 - GA-EP45T-UD3R (rev. 1.0) if you need to download the latest ET6 version.
The latest version is B12.1122.1, size = 42.59 MB and date = 2012/11/22.
The only time that I've seen different ET6 voltages is with Reference voltages.

Raise the CPU Vcore to 1.28750 and increase Cpu Termination to 1.24 volts and rerun Prime95.

BSOD 0x3b
Have you recently installed new device drivers in your system? If yes this might be part of the problem.
In http://answers.microsoft.com/en-us/w...3-1fda74a7315e, a recommendation near the top of the page says:

Try removing and reseating all cards, memory, and cables (both ends when possible) - actually
remove and replace - do not just snug (on a PC) and clean out the dust bunnies and vents.

Always unplug the power supply's A.C. power cable from the wall outlet before you remove or install any hardware components.

[JGPliskin]

Re: How to Overclock Q9550 on EP45T-UD3R with DDR3-1600

No sir, The ET6 is not running when i test my system.
ET6 is on the latest version sir.
I reseated my rams and gpu and raised the vcore to 1.28750v and the cpu termination to 1.24v, i still got a BSOD, but the cause is different now.

[profJim]

Re: How to Overclock Q9550 on EP45T-UD3R with DDR3-1600

In post #11 I said

Make the following changes, but make one change at a time and then test again.

This means that you only change one setting at a time and then you run Prime95. It the first test had errors, then you would change the next bios setting and test again. If there were errors each time, you would have performed four or more tests. In post #14 it looks like you made changes to all four bios settings and you only posted one screenshot and it shows four bios changes.

In post #17 you posted the screenshot for your BSOD 0x24, but you didn't say if it happened while booting, or if you were running Prime95.

Do a web search using: bsod 0x24 windows 7 and you will have links to forums that deal with the specific bsod.
The sevenforums.com forum is one of the best sites and their information is excellent.
note: bsod 0x24 is shorthand for bsod 0x00000024 where 0x means that the following eight (or 2 or 3) digits represent a hexadecimal number code.

You should go back to your stable 366MHz settings and the only change you should make is to manually enter 11-11-11-28 primary memory timings.
When you boot your system with these stable settings, run the Windows Command Processor as an administrator and run each of the following commands to see if there are any NTFS problems with your boot partition:

• chkdsk /f
  
• sfc /scannow

It's important that you include the HWiNfO sensor screenshot so that we can check voltages, temperatures and memory info, and you need to provide the Cpu-z, Memset and ET6 information. Luke did an excellent job posting detailed information over and over again throughout his Socket 771 Mod Xeon thread.

I'm done asking for detailed information, if you don't provide it each time you complete a test, I'm done helping in your thread.

If there are no errors the best way to proceed is to by splitting the difference with your bus speed.
366MHz was stable and 402MHz has problems, split the difference by changing bus speed from 366 to 382MHz (16MHz is half the difference) and then rerun your Prime95 tests. If there are no errors, you can split the difference again by increasing the bus speed from 382 to 392MHz (10MHz higher is half the distance between 382 and 402. If the 382MHz test has errors, you can try my suggested bios changes where you only change one bios setting at a time and then repeat the tests. If these tests all have errors, then you split the difference between 382MHz and 366MHz and start over at 374MHz.

There are many reasons why you might not be able to run your 4 x 4GB memory kit at its 1600MHz rated speed
Right now your stable 1464MHz settings are 91.5% of the 1600MHz memory speed.