Specs: CPU: Intel Core 2 QUAD Q6600 Quad Core 2.4GHz 8MB(2x4MB) FSB1066 GPU1: XFX GeForce 8800GTS 320MB GDDR3 GPU2: EVGA GeForce 8500GT 256MB DDR2 RAM: Corsair TWIN2X2048-6400C4DHX DHX 2048MB/2GB x2 (4x1024MB) MOBO: Asus P5N32-E SLI PLUS nForce Dual X16 SLI Soundcard: Sound Blaster Audigy 2 ZS PSU: Hiper HPU-4S730 Type-R Omnigrid II 730W ATX(BTX) CPU Cooler: Zalman CNPS 9700 GPU1 Cooler: Zalman VF1000 LED Ultra Quiet VGA Cooler VRAM1 Cooler: Zalman ZM-RHS88 VGA RAM Heatsink GPU2 Cooler: Zalman VF900-CU GPU Cooler Drives: 500GB Western Digital Caviar SE 16MB SATA II 500GB Seagate Barracuda 7200.10 6MB SATA II 320GB WD SATA 200GB WD SATA 200GB WD IDE 160GB Seagate IDE Optical: NEC DVD/CD +/- Burner on IDE -> USB2.0 dongle Case: Nzxt Lexa Aluminium Case 3x 120mm fans 2x 80mm fans, Glossy Piano Black Finish OS: Vista Ultimate 64Bit (Being on MSDN got it's advantages) All temperatures I can measure (2xGPU, 4xCore, Mobo & Drives) are all fine, sub 50C and I got beyond overkill amount of case fans and they're all set to full tilt so the temperatures are not the issue here I'm fairly certain. With normal use (Firefox, Photoshop, VLC etc) I can run the PC indefinitely with a stable feel and good performance, gaming feels the same right up to the moment the computer reboots, no BSoD, no hickups near the end, nothing, just a plain reboot, same as if I hit the reset button. Happens in World of Warcraft, Unreal Tournament 3 Crysis Gears of War etc. Steps I've taken: I've gone through the cooling thoroughly, made sure all heat sinks are properly seated with just enough paste to mitigate any imperfections in the surfaces, all temperature readouts in applications like speedfan are fine. Removed 2 ram pieces then swapped after it crashed like that also, didn't help. I've tried gaming on my other video card the much weaker one I only use to power 2 screens, didn't stop it from crashing. I don't have another CPU that fits in a 775 slot so I haven't tested if it's the problem. I've fiddled with the bios settings but I don't dare do too much in there, I went to the official Cursair boards and made sure my ram settings were correct (which they were). I used to run the system a bit overclocked but during the course of my investigation I've set it all to default speeds without any improvements. All drivers are the latest I could find. Any suggestions are greatly appreciated.
tell me, you're not running those two GPUs in parallel (cause you say - both my GPUs and 4xcore temps are OK...) anyways, maybe you should look into your BIOS settings - helped me with the very same problem. First things first - go and check out the PCI-x settings wether there are no 'accelerating' features enabled. Mine had some double rate or something enabled which didn't handle anything well and crashed when gaming. Also, try and check out the clocks for your RAM. I had two DDR 256 mb sticks which were supposed to run at 400mhz; guess what - they didn't. When I left only one of the sticks and pulled the other out, everything was as stable as you could wish. (so, that's another thing - try and leave only one stick in the mobo, maybe you'll see some changes and then it'll be clear where to search for the answer). So then what I did was, I lowered the mhz to 333 and everything was 100% stable. no restarts since. Later, I switched to 2x1gb 533mhz sticks, but these work perfectly. So, those DDRs had some problems. hope that helps!
I use both video cards since I got 4 screens, the there are no acceleration options associated with PCI-X in my settings but I can change the clocks of the 3 PCI-E x16 slots on the board, as well as "SPP <-> MCP Ref Clock" I got it all set to 100 and auto (defaults) I can also enable "Spread Spectrum Control" of a number of items like CPU, PCIE, MCP PCIE, SATA & LDT at this point my eyes start drifting together... I can also enable/disable: Intel SpeedStep CPU Internal Thermal Control Limit CPUID MaxVal Enhanced C1 (CIE) Execute Disabled Bit Visualization Technology LDT Frequency Which I don't have a clue what they mean either, I run on defaults. I'll try the ram thing, but I went to the official corsair board and made sure my timings and Hz are all correct, and they are. :|
My first guesses would be RAM or PSU. Both of those can cause random lockups during high system activity. The RAM's easy enough to test, just run Memtest86+ overnight. If your system has any errors during the tests, it's time to start pulling RAM until you can eliminate the problem. If after running memtest overnight you do discover you have some bad RAM, you can perform a high-utilization test similar to a gaming experience by choosing test 8 and letting it run repeatedly. That will usually throw errors very quickly on bad RAM.
