So you're telling me that 16-bit real mode still has a freaking use?
Yes. The bios uses it. It has to, to be able to start DOS. (Yes, really, it's still backwards compatible).
It doesn't, just takes up space on the die.
Not much.
Protected mode is pretty much deprecated at this point as no major OS for x86 is 32-bit only,
Huh? Windows 7 32-bit? Sure that all exists. For small devices, tablets etc, why do you need 64 bit addressing?
so that leaves long mode, which you can't, as far as my coding has gone, escalate to without passing through real and protected. Seems pretty dumb.
It depends on what you call "dumb". It allowed a migration to the state of today without breaking compatibility. And this "compatibility" is the key issue because it allowed people to re-use their hardware, at least for a while. It is an important issue if you want to keep your market intact. Hence, it might look pretty ugly from an engineering perspective, but from a market perspective it makes complete sense.
68k at least has the saving grace of being big-endian, and since 68k has no 64-bit extension... I'm a little bit confused as to why even bring 68k up on an article about POWER.
Little endian, big endian... it's only a convention. Nothing to bother about. POWER is pretty much irrelevant in the desktop market though.
Also as I said, I'm against the patents on the AMD64/x64 architecture, as well as the wide use of proprietary and glitchy EFI/BIOS, as well as crap like SecureBoot ( which is also on Windows ARM rubbish
Bios is not really proprietary anymore, but its hardly a good basis anymore. UEFI is just another attempt of Microsoft to regain control over the market -quite obviously. They don't want to sponsor cheap tablets and allow people to install the Os of a competitor on it. Understandable from a market perspective, but I really wonder why no consumer organization protested against this obvious protectionsm of the market.
OpenFirmware, which is used on most POWER devices, is open source, and with OpenFirmware, any bugs or limitations are publicly documented. Not saying it's perfect, as I don't like it's Forth-style syntax for command shell use, but it is a better alternative to proprietary UEFI.
Not really. OpenFirmware is close to unsuable, and it's also mostly marketing nonsense - OF only works half as good as it should, if it works at all. I've here an old G3 Mac with "OpenFirmware", which is a complete piece of crap implementation. It is so buggy, it can only boot MacOs, or something that looks very close to MacOs. Tried over a year to make OF boot Linux without success, one bug after another, nobody tested this stuff, and I finally got it working by a "fake MacOs" OF was able to boot.
The main reason I put this out was because, most POWER equipment that is new, is thousands of dollars higher than comparable x86 equipment, and A-EON or A-Cube partnering with TYAN for manufacturing a board based on the specification may serve to reduce production cost. The main issue for A-EON I can see is lack of a clear upgrade path from the 10-year old PA6T CPU, so unless they can pull a design out of their own arse, I don't see any better option than to partner with TYAN.
POWER addresses supercomputing markets, not average desktop markets. The desktop branch (PPC) died years ago, the same death as 68K died before - lack of a proper market, and a customer base that was too small to make it sustainable. x86 dominates the desktop markets, only endangered by the ARM underdog which attacks it from the low-end. "Cheap, lean, low-power", hence, the dominant species in the mobile devices market - which is exactly the market M$ wants to gain access to (and - as it seems - hopefully fails to) and has invented UEFI to protect it.
