I know this is a really old thread, and I hate to be the guy who performs internet necromancy on something that's out-dated, but it's so far the best resource I've found for answers to some questions.
One thing I'd like to ask the author of this excellent guide, if you're still paying any attention to this thread, is just how effective you think this method would be on a 32-bit Windows XP machine?
It's a bit of a pet project of mine, just something I'm toying around with to see what kinds of performance I can coax out of a Frankenstein machine on as little of a budget as possible. It's based on a stripped-out HP Pavilion a1630n. I put a new HDD in it and installed the machine's original factory OS, which was WinXP MCE 2005, and I'm about to see how far I can push the stock Asus motherboard.
I'm starting with this motherboard's HP-approved maximum - 4GB of PC5200 RAM - but I've obviously run into the problem of 32-bit Windows only having access to 3.2GB, which leaves me about a gig shy of what I've got.
I read on another site (I can't post a link since I'm a new member - a google search for "Debunking the widespread myth of 2^32=4GB" will bring up the article), that "patching the kernel won't help Windows XP see more than 4GB of RAM". Does this mean using the method in your guide won't help at all, or that I can get the benefit of this guide's method to get XP to read all 4 GB of RAM, but no more than that?
If it's the latter, I'm okay with that, and I'll just forge ahead knowing that 4GB is all I can get with this machine without upgrading the OS (which I may do). But if it's the former, I'll definitely need to upgrade.