I work with a lot of different consumer hardware, and I run into many problems with said hardware. I am attempting to document problems (and misbehavior/invalid specifications) that I have run into with various products. Most of these issues have not be resolved by the manufacturers in question.

ASUS M3A78 Pro

• Manual and site specify that the maximum CPU voltage is 1.7V, in actuality the maximum is only 1.3V (this is fixed in a test BIOS version 0902 which has not been released to the public, you can obtain it here).
• POST screen will hang while loading NVRAM if a USB device is selected in boot device order (fixed in BIOS 0801 and greater).

ASUS M2N-SLI Deluxe

• Linux compatibility problems with USB (and other random freezes) using BIOS past revision 1502.
• Problems booting Windows on newer BIOS, just plain broken.

ASUS M4A79T Deluxe

• Instability with BIOS past revision 1103. Problems may show up as memory errors, etc. Downgrade to BIOS 1103 for stability.

Gigabyte GA-EP45-UD3P

• Manual and site specify that the mainboard supports 16GB RAM (4x4GB DDR2-800), but in real world (memtest86+) tests system is unstable.

HighPoint RocketRAID 3510

• When RAID controller initializes, it may not detect all hard drives even if they are all spun up thus creating broken RAID arrays.
• Enabling "Staggered spin-up" on Western Digital WD5000AAKS (and maybe others) may cause the drives to be unable to be used by any SATA controller besides Highpoint due to a hard drive firmware bug.
• While Highpoint claims to support FreeBSD, the management tools do not work in FreeBSD 7.1.
• Controller sometimes causes machine to hard lock, will not be recognized on boot until you leave the computer off for 10-15 seconds.

Supermicro X8DAi

• Any BIOS past revision 1.0a does not support CPU throttling under VMware 4.0 and above.
• If you are using a BIOS past revision 1.0b causes issues with 3Ware 9650SE. Symptoms may be the system refusing to go past the POST screen, freezing at 'storage-drivers' on VMware ESX startup.
• While Intel advertises the Intel ICH10R chipset as supporting PMP, this is incorrect. Supermicro has confirmed that this is a typo on the specification sheet.