And the RAM has been fine in heavy use yesterday in both Windows and OS X.
My 4 gigs that failed the memtest86 tested successfully for hours yesterday using the Windows version. I found no mention of it anywhere having a problem with more than 2 gigs, but it's the only conclusion I can come to. I think that memtest86 can't do 4 gigs of RAM at a time. I replaced the mobo with a brand new one and same thing happens. Individually, the RAM passes, in pairs, it freezes memtest86. All HDDs pass their SMART tests and manufacturer diagnostics. System is very well cooled (reports nice, low temperatures, in its Antec P180 case), and everything except for the PSU, a leftover from the last build, has lived every second of its life on a line-conditioning UPS, in a building with nice clean power. If I'm not lucky, I'm going to lose my ever-loving mind. Maybe, if I'm lucky, it'll be a bad stick of RAM, and I can just RMA it to Corsair. where do I go from here? Bad motherboard? Bad (Antec NeoHE 500) PSU? The system has done this from day 1, when it just had two sticks of RAM, rather than the four it has now. I have two more sticks to go, but assuming they both test clean. Memtest completes an entire pass successfully. In frustration, I strip it down to just a single stick of RAM (and all the HDDs have long-since been disconnected). I drop into the BIOS and force it to take the proper RAM settings (5-5-5-18, PC2-6400, 1.80v) in case the Asus P5Q-E can't automatically figure things out. So I pop out two sticks, and Memtest now reboots when it hits test 7. With all the RAM installed (4 2GB sticks), Memtest86 freezes five seconds into the first test. The chances of two bad HDDs (the spare being brand-new) is pretty rare, so I wander off to run Memtest86.
This is a known-good install DVD, so I try again. On first boot after install, Windows tells me that explorer.exe is not a 64-bit compatible application, and then freezes up. So in went the spare 80GB drive, and I installed a fresh copy of Vista Ultimate SP1 64-bit. Maybe the problem was the HDD, and not the video card? This merited testing. Well, the Raptor replaced one that had crib death, so reliability was not to be expected, and it had been making (head reset?) clicks during the video driver's non-BSOD incidents. So a new video card was ordered, and while I was digging through Vista's Event Viewer to get the exact error message, I stumbled across another error (OS-drive file access failure) that Vista flat-out told me was due to failing hardware. That something being replace the video card. Then it started happening again, and I decided that it was time to do something about it. Changing versions of drivers made the issue much less frequent, so that seemed to make sense. At first, I blamed the regular video driver crash/recoveries on ATI's long tradition of awesome hardware and crap drivers. This build has been troubled since day one.
0 kommentar(er)
