Order of testing/replacing to solve MemTest86+ failures

gocard

Distinguished
Oct 23, 2011
1
0
18,510
Hi,

About a week ago, I replaced my motherboard, cpu, and ram in my system and did a fresh install of Windows 7. I'm now running:

Intel Core i5 3750k (new)
MSI Z77-G41 Motherboard (new)
G Skill Ripjaw 32 (4x8) GB DDR3 1600 RAM (new)
Sapphire Radeon 7770 1GB (existing)
Nexus Value 430W PSU (existing)
Crucial M4 128GB HD (existing)

After setting everything up, and it worked flawlessly for the most part, except if I plugged in 5 USB devices, the usb connection to my UPS failed. I did some gaming on the computer, and used Netflix and stuff, and everything worked just fine.

I monitored heat pretty closely after setup, and everything was fine. CPU average was 35C during idle and low 50s under stress. Video Card was 30C idle and mid to upper 50s under stress.

Just today, on booting up the computer, I suddenly started getting BSOD. I ran MemTest86+, and immediately a failure is listed. So I pulled out 2 of the RAM sticks and reran MemTest86+. I let it run 2 of the tests on all 16GB and no failures (so I let it run about 40% of the tests), so I thought I had isolated the problem. I boot into Windows, and it seems ok initially. Then I start getting crashes again. I rerun MemTest86+ and almost instantly, I get a ton of memory failures showing up.

My question is, what are the likely culprits, and what should I be testing next? I guess instinctually, I'm wondering if my USB issues and now RAM issues indicate that my PSU is failing or underpowering the system. Or is it likely that multiple sticks of RAM are failing? Or is it a motherboard issue? Or could it even be a CPU issue?

Any advice you can offer, or if it were you, what's the order of replacing and testing that you would do?