ASUS Maximus Formula SE

[ssideratos]

Look at this post
http://www.xtremesystems.org/forums/...postcount=1678

OCCT is being rewritten. It is mistaken about, because of that "30mn PC Stable" message. New releases will put the minimal test at 1h. And never forget, to be 100%, you'll need a few hours run like any other proggy.

In RAM mode, OCCT is much more sensible than Prime95 Blend

If only OCCT fails, just run it at stock. If it doesn't fail, you know the answer: The PC is tsable until you find a game/program that stresses the RAM/NB subsystem and you'll hang up. Also, if you have an unstable PC, each time you have bugs, you'll have to test without OC to rule out an unstability. So, it's better to run a rock stable config, rather than a so so stable

I've spent two weeks nearly full-time trying combinations and stability testing and have yet to reach a satisfactory state. Maybe someone could give me some advice?

Key Config Points
Corsair 620HX
Q6600 G0
Thermalright Ultra-120 xtreme with AS 5
2 X 1GB PC Ballistix PC-8500 Tracers (and another set non-tracers)
Sapphire Radeon HD 2600XT 512MB
WD 150GB Raptor HD

I'm on my second board -- The first would either not boot at all (all lights on motherboard lit up) or fail while in BIOS (no operating system ever installed).

NewEgg made RMA fairly easy.

On this second motherboard, I've gotten much further and installed Vista Ultimate 32-bit, but not until after about a million power cycles. Even minimal changes to BIOS from stock settings require me to cycle the power not only after the change, but in order to POST after each full shut down from Vista.

Stability is very important, as such I am only looking for a very modest overclock (3 GHZ Max). However, there are very limited changes that I could make that didn't require me to power cycle after each shutdown before I could boot again.

I've resigned to the fact that this motherboard may require a power cycle after most BIOS setting changes, however, I find it unacceptable to need to reach to the back of my PC and cycle the power each time I want to boot.

The limited overclocked BIOS settings that did not exhibit this behaivior included CPU Level Up to QX6850 giving me my 3Ghz with the MB Auto Volting the CPU at 1.34-1.36. On Auto, RAM however was set at DDR-800 with 5-5-5-18-42 and vDimm of 2.0v as reported by PC Probe. The RAM seemed way to conservative for the Ballistix and my bandwith and latency settings were pretty bad.

Switching to manual and changing RAM to DDR-1066 at with Auto, or any timings require the power cycle.

At DDR-800, I am able to make some changes without encountering the Power Cycle required issue including reducing the timings to down to as low as 4-4-4-10-25 while remaining memtest and Prime95 Blend 8-10 hour stable. (vDimm remains at Auto - reporting 2.00v by PC Probe).

I was willing to settle for this when I tried OCCT RAM mode for final stability testing. It consistently fails between 28 and 38 minutes.

I tried to give the vDimm more volts, but whenever I set vDimm at 2.10 (2.20 real) or above in the BIOS the system requires power cycle after each shutdown. 2.08 equal to 2.18 real and OCCT still fails in the same time range.

After 100's more combinations and cycles, I decided to test OCCT at stock.
OCCT still fails at stock 2.4Ghz with DDR-800 5-5-5-18. With any combination of the Ballistix.

I've even tried going back to 0505 BIOS from 0701, but my results are even worse there.

The system is consistently autovolting the NB at 1.46. Any change and there is either a failure to post or the power cycle issue comes back.

I have read every post in this forum and can elements of the issues I'm facing, however most seem to be encountering these issues at overclocks of 3.6 and above.

I am about to crawl up into the fetal position and cry or get medieval on this board. Over two weeks of nearly full-time troubleshooting and testing tyring to achieve OCCT stability and I'm ready to trash this thing and get a Dell.

Does anyone have any ideas before I do that? Brief testing of the CPU and RAM combinations before OS installation shows it to be at least memtest stable even at 3.4 Ghz and DDR2-1150 on air (I didn't try higher than that) but the power cycle issue is there after OS installation. Could it be the motherboard again? I don't know if I can endure the pain again of another RMA, installation and testing cycle(s).

If I have to do that, I think I'll just use Prime95 stable at 3.0Ghz, DDR2-800, 4-4-4-10-25 settings for a little while until maybe a new BIOS does better for me, or if I notice any real world instability at all I'll figure out what replacement MB or Dell to get next. (OCCT isn't stable even at stock).