Linus Torvalds on 5/13/07 wrote:
---------------------------
>[...]So no, PAE does not mean that you can use more than
>4GB of RAM. Even before PAE, the practical limit was around
>1GB, and PAE didn't move that post a fraction of an inch!

Can you please elaborate on the practical 1 GB limit a little further ?

(Lets leave PAE aside - I think you are being too harsh on the hw designers; I doubt that any engineer ever thought PAE was anything more than a temporary hack, as was HIMEM actually, but it was probably much easier/cheaper than designing a 64-bit CPU. What marketing did with that is another story ...)

Do you mean that it is absolutely impossible for a 32-bit OS to efficiently use RAM > 1 GB, or that it complicates the kernel compared to a hypothetical scenario with the same amount of RAM and 64-bits ?

I understand why it is extremely convenient to be able to map the entire RAM entirely in kernel, with plenty of space to spare, however this hasn't been the case for some time. So, presumably all the complications are already there and have been taken care of, and are being maintained, etc. They are not going to disappear in the next 10 years (while Linux is still being used on 32-bit embedded systems).

So, in practical terms, is there benefit from using 64-bit OS on a machine with, lets say, 2-3 GB RAM ? (Please, note that I am not arguing a point - just asking)

regards,
Tzvetan