You must log in or # to comment.
Itanium was an interesting architecture but it relied on compilers to build efficient code for it’s bundles. This compares to x86 which dedicates loads of silicon to getting the best performance even out of mediocre code. Unfortunately the Itanic architecture was really ill suited to emulating the other.
ETA: rm dups.
It dedicates dedicated silicon?
All the cache and prediction logic which eventually gave us spectre is basically compensating for whatever crap random compilers kick out.
I was only making a joke, and not asking for clarification. Do you not see the redundancy of saying “dedicates dedicated”? I thought my reply would have made it obvious.
