High-End System/Mediocre Power Supply?

[Mr. SlicknHard]

Ever since I add a second 7900GTX, I have been experiencing random reboots at the Windows login screen. The reboots usually occur when I first turn on my system, generally not after a restart, but they are random. Also, when I exit games like Oblivion and Fear, my system will hang on occasion. Could this be a power supply problem?

My system specs are as follows:

DFI LAN Party Expert
AMD 4000+
2GB Corsair 3200XL Ram
4, 300GB Seagate Barracudas in Raid 0
Creative XFi
2, BFG 7900GTXOC Graphics Cards in SLI
Antec 550W TruePower 2.0, with EPS12V

One more thing, when my system hangs or restarts, I never receive a Windows error. So I have ruled out any type of OS issue.

Thank you in advance for your comment/advise. Cheers!

 

[zandor]

It's hard to say really, but this sounds more like a driver/software issue. What suggests a software problem to me is when the reboots happen. If it were a PSU issue I'd think you'd be much more likely to boot sucessfully and then crash in the middle of a game when PSU load is at it's peak. When you're booting the machine the video cards aren't doing much of anything.
I played through Oblivion on a single 7900GTX, and noticed I frequently had crashes on exit. In my case though it was only the game that crashed- just got the 'ol report crash to MS dialog box. I wouldn't be a bit suprised if adding SLI tickled something in a driver that takes down the whole rig when that happens.
A Windows error is not at all guaranteed when a software crash happens, particularly if it's driver related. Drivers have direct access to hardware. If a driver screws up, it can wipe out the whole system causing a hang or spontaneous reboot with no error message at all. Apparent system hangs can also be caused by the graphics processor locking up. If that happens Windows may still be running on the CPU, but is unable to display anything since the vid card is hosed. This used to happen to me all the time on Linux back when I was using a "semi-supported" video card. The screen froze, but I could telnet into the machine from another system and do a clean reboot or kill the X server.
Windows tries to detect and report such errors, but it's "best effort" (or maybe half assed... you'd have to ask an MS engineer) reporting.
I'd check around for driver updates or try some older drivers.
You might also be able to alleviate the start up problem by messing around with the bios. It's kind of a long shot, but slowing down the boot process (turn off fast post, maybe turn on the memory check) might help if it's crashing because the vid cards need more time to get ready to do some work. Mostly I'm just suggesting this since I can't think of any good reason it should crash on a cold boot but not on a warm boot. Usually it's the other way around.
All that said I'd probably upgrade the PSU anyway. Antec's decent but for your setup it's too close to "minimum" to make me nice and comfy. Call me paranoid if you like, but I just like to have more margin.