Re: Is Intel exceptionally unsuccessful as an architecture designer?

Even if 99% is correct, there were still 6-7 figures worth of
dual-processor x86 systems sold each year and starting from 1997 at
least tens of thousands of quads.
Absence of ordering definitions should have been a problem for a lot of
people. But somehow, it was not.

My guess:

- Luck due to the relatively limited amount of reordering taking place
  in the CPUs of the time.
- Limited software support, encouraging very coarse commmunication
  patterns (like parallel `make` or processes communicating via pipes)?
- The remaining cases were sufficiently rare that the victims blamed it
  on themselves for pushing the boundaries (and added hacks to work
  around the problems instead of complaining to their CPU manufacturer
  about the insane semantics imposed by their hardware)?


        Stefan