Note that this doesn't enable you to run 64-bit applications, or for any single process to address more than 4GB of memory.
As a result, it's of extremely limited utility today - you'd need a 32-bit application that needs to run on bare hardware (ie, not a 32-bit OS running under a 64-bit VM), and needs a ton of memory, or a pre-64-bit system with > 4GB of memory, which would be both rare and inefficient, as that would be single core/socket systems that are at least 6 years old at this point.
I guess if you had some system you were trying to keep alive that needed a lot of RAM, PCI-X card slots, and 32-bit software it might be handy, but that seems very rare/esoteric at this point.
A wild guess could be that the work needed to be done on that station was windows only or they did not had *nix people on board or that they wanted to see if this was enough ram for Crysis ...
The developer had lots of Windows desktop experience but little Linux desktop experience. We run a mix of Windows, Mac and Linux here, so using Windows shouldn't have been a problem.
Long hardware maintenance cycles, especially in businesses.
The last 32-bit NetBurst (Pentium 4) systems with 2GB of RAM aren't that bad to work on, when given a video card that supports Windows 7/8 video features and a decently fast disk.
You don't need a whole lot more to run a greenscreen terminal emulator or a basic web browser.
I'm actually in the market for a old dumb box and for power(i.e. energy/electricity) reasons I'm shying away from the PressHot - Pentiums. I'm targeting a Core 2 Duo which seem like a great balance between power and performance and I'm reasonably sure that the money I'd save by going this route would not be recovered over the course of 5 years if I get a more recent box.
Unfortunately Win8 requires NX so only the very last ones will work and Intel released socket compatible 64-bit capable NetBurst CPUs not long afterwards.
Sure. I have a scanner that is now useless because Canon never released a 64-bit driver. (The useful lifetime of a scanner is longer than the life of an individual PC or the time a manufacturer will support it.)
Lots of scientific and industrial equipment will not have 64-bit drivers.
You could use it in a 32-bit VM. Since I went 64-bit, I've been using my Lide20 CanonScan though a 32-bit XP install in VirtualBox. Sure, the VM takes 20 seconds to boot to desktop, but that's better than junking a working piece of hardware simply because the manufacturer doesn't support it.
1) 64-bit has bigger overhead. If you have, for example, 8 GB maximum (e.g. a notebook) then you'll have the advantage of having more RAM available by using 32-bit system. If you just use a notebook for surfing, you wouldn't care. If you did something that had specific memory needs, it was the best solution before the notebooks with more RAM possible appeared.
2) Other hardware compatibility as there is still enough hardware without 64-bit drivers. But don't expect too much there: some kind of such hardware will have problem with this patch though. You'd have to try to be sure.
Now that even notebooks allow 16 GB the patch is becoming less relevant.
1/If you have 8 GiB of RAM, and you just surf, you don't care about the 64-bit overhead. Also 64-bit is more convenient as browser tend to flirt with the 2/3 GiB process limit.
2/If you have 64-bit machine, this problem is unlikely to exist on Windows.
The Geoff Chappell article is great. Seems like he examined it in detail:
"The 32-bit editions of Windows Vista and Windows 7 all contain code for using physical memory above 4GB. Microsoft just doesn’t license you to use that code."
PAE was turned on but the licensing control code made by Microsoft checked if it's a consumer version of the OS or if it's a server, and if it's a server which version, and based on this decided how many memory to give to 32-bit OS. Consumer versions remained limited to 4 GB. This is detailed in the analysis of Geoff Chappell:
I'm pretty sure that not all the drivers(especially graphic ones) fully support PAE. Server versions of PAE would be much easier to support than making sure drivers for win32 would be compatible in a PAE environment.
Typically, device drivers must be modified in a number of small ways. Although the actual code changes may be small, they can be difficult. This is because when not using PAE memory addressing, it is possible for a device driver to assume that physical addresses and 32-bit virtual address limits are identical. PAE memory makes this assumption untrue.
How safe is it to use? Have you heard of people having issues with the patch? If things don't work, would it be as simple as removing the patch or would a fresh Windows install be required?
As a result, it's of extremely limited utility today - you'd need a 32-bit application that needs to run on bare hardware (ie, not a 32-bit OS running under a 64-bit VM), and needs a ton of memory, or a pre-64-bit system with > 4GB of memory, which would be both rare and inefficient, as that would be single core/socket systems that are at least 6 years old at this point.
I guess if you had some system you were trying to keep alive that needed a lot of RAM, PCI-X card slots, and 32-bit software it might be handy, but that seems very rare/esoteric at this point.