A glimmer of hope amidst silicon carnage and despair: XP made it through NTFS formatting and is now installing the OS. Before, it crashed and burned just uncompressing and writing sys files to the HDD -- never reached installation and configuration phase.

Got XP installed. So far so good.

Hey, hey ...

Awesome. So moving the memory to slots 2 and 4 seemed to have done it.

I know that on the P35 boards slots 2 and 4 (black ones) are the best for running the memory at higher speeds, did wonders on my P5K deluxe...

I'm starting to have doubts about this RAM tho, no way should you need 2.0 vdimm with the RAM running at 1066mhz and 7-7-7-x timings to install the OS.. keep us posted

It's not the RAM - passed mem-*test*-('"). The la-*test*-('") is that Jason will probably RMA the board for a replacement, but will try a different cpu first.

Sorry, that'll teach me from just browsing the posts instead of reading them.. Hope the board hasnt damaged the RAM either...

I swapped the CPU. Works fine in my other box. Must be the mobo. I put in an order for a new one tonight.

Fat Elvis checking in . . .

A QX9650 Extreme Edition CPU and new Maximus Extreme mobo arrived yesterday. I decided to order a second mobo and RMA the 'Three Mile Island' board, given that FW is around the corner. Never a bad idea to have a spare on hand.

The native NB/SB/Vreg cooling on the Maximus Extreme is a nice feature and works well once it stops leaking. lol. Actually, not very funny. Why, oh why Asus chose to use 3/8" barbs with funky 1/2" converters and not go with standard 1/2" barbs eludes me (unless they're selling more motherboards in Europe than in the US - doubtful). To have to step down a 1/2" loop to 3/8" impacts a loop's efficiency.

After installing the new mobo, the system booted up, as before, but abruptly powered down and wouldn't reboot once I was in BIOS. I unplugged the computer and after 20 minutes or so, it booted back up and has been less mercurial since. Seems to be running fine now -- for how long, who knows?

One theory: I confirmed through Everest UE and Cpuz that when manually setting vdimm, the BIOS setting reads .08v lower than the actual setting. For example, the RAM I'm using, Super Talent DDR3 1866 MHz, runs at a recommended 1.9v. I have the vdimm in the BIOS set to 1.82 in order for the RAM to run at 1.9v.

One of the problems with the other board and perhaps the reason the computer suddenly powered off today with the new mobo installed is that vdimm voltage was unknowingly set too high since the board overvolts by .08v.

A few interior shots with second Maximum Extreme installed -- notice native wcooled blue NB heatsink:







With 12 120mm fans (three exhaust on top of case; one exhaust on back, small rad; three on each of two triple radiators; one on HDD caddy; one side panel), three fan controllers are required. Now you know why I dread swapping out motherboards -- the wiring is ummm a challenge:

BigNG fan controller mounted on top of PC P&C 1KW PSU:





Second BigNG controller and MiniNG controller mounted on back side of case. You can also see the back side of the two triple rad's here -- mounted one on top of the other:



Three exhaust fans mounted on top of case. A fifth RaptorX drive is mounted east of the fan panel in a Lucite enclosure so that it's visible from the case top exterior:

That is bizarre behavior from a mobo. Great catch BTW, I would've never looked for a Vdimm Voltage drop. Something new for all our troubleshhoting kits. Thanks!

It's a vdimm voltage hike not drop, isn't it?

Yes, as odd as that sounds -- correct. I'll take a few photos of the BIOS setting and Everest readout to illustrate.

Pics: "Every picture tells a story, don't it?"

Pretty sure I read about the .08v vdimm boost in one of the articles about 1800Mhz ddr3 in which they were using your (our) board.

You're running the QX9650 now, Jason?

Just to emphasize -- it's an overvolt, not a voltage droop -- a .08 voltage spike.

Beyond the obvious stability issues, I think the thing most significant about this mis-calibration is that DDR3 is very easily damaged when running at a higher than recommended voltage. DDR3 is apparently easier to burn out than DDR2 when stressed at voltages higher than mft specs. This according to engineers (not a sales rep; engineers) at US made Super Talent RAM.

Yes, sir: the QX9650. Will have to change the signature!

Oops, my bad. It was late when i was reading that.