|
|
|
|
|
by aidenn0
1618 days ago
|
|
The Intel C compiler was already well established as a top IA-32 compiler by the late 90s (prior to any IA-64 release). This article[1] from 1999 assumes Intel is responsible for the compiler. My recollection is that the primary focus on 3rd party software was getting systems software ported. I don't think Intel was banking on 3rd parties making compilers. A lot of 32-bit architectures not named "68000" from the 80s/early 90s suffered from poor first-party compilers and a lack of good 3rd party compiler support; in 1980 an optimizing compiler was not considered an important part of a microprocesor's ecosystem, but by the time IA-64 came around the importance was fairly well understood by hardware vendors. Given the quality of the first-party IA-32 compilers, I think Intel (and everyone else) expected that the first-party IA-64 compilers would be good. Certainly by the release of Merced (and likely well before), compiler engineers internal to Intel were aware of how hard it was to codegen for IA-64. Certainly during the time period that Intel was pushing IA-64, they had an insatiable desire for compiler developers with advanced degrees. 1: https://www.cnet.com/news/intels-merced-chip-may-slip-furthe... |
|