On 3 June 2013 23:33, Andrew Hoerter <amh at pobox.com> wrote:
I'd guess one major factor is that Intel's early market domination
allowed it to sink lots of money into chip design and R&D -- IOW, they
succeeded in spite of the architecture, not because of it. "With
sufficient thrust, pigs can fly."
Yes, I can't really disagree with that.
Except perhaps with the slight cavil that Sun and IBM, in particular,
also threw really quite a lot of money and effort at their RISC lines,
and HP and Intel at its non-traditional line, and still failed.
Indeed, if I were to pick one single move that has saved x86 and kept
it relevant today, it was not Intel at all, it was AMD and the
Sledgehammer architecture. Intel would not have done that, because it
was still hoping for Itanium to win.
When it became apparent that Intel was going to have to do a 64-bit
x86 chip, it did its own proprietary one. Microsoft vetoed this,
saying that no, it already supported the existing incompatible (and --
subtext -- crappy) 64-bit CPU, and as MS was already supporting AMD's
really rather good 64-bit extensions, Intel had better be compatible
with AMD and support the same extensions too, or Intel would find its
shiny new instruction set unsupported in Windows.
3rd party reports are that this was /not/ a happy, tranquil, relaxed
exchange of views.
But Intel caved, and licenced AMD's ISA - which must have stung its
pride more than I can readily imagine.
--
Liam Proven ? Profile:
http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at
hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884