@gloriouscow Do you know if the meaning of 'intel microcodes' in 1976 would match the contemporary one; or would it have been much less clear at the time whether they were talking about what we mean by 'microcode' or whether that was more of an ISA license?
That conceptual distinction would definitely have come early enough; but I don't know what the state of the terminology was.
@fuzzyfuzzyfungus i'm trying to find the actual verbiage of the 1976 patent agreement
here we go:
https://www.ardent-tool.com/CPU/docs/MPR/19920708/060901.pdf
seems they would have saved themselves a lot of trouble by adding "...Written before 1977" to that clause
another possibility is that they actually did intend to grant AMD a perpetual license in the first place but then tried to weasel out of it.
AMD was an official second-source manufacturer for the 8088 and 80286 so it clearly wasn't a problem for those particular chips
@gloriouscow I wonder if there were some real iAPX 432 optimists involved.
The contract still seems weirdly vague; but that seems like the least baffling explanation for why they might have seen a fairly broad license to AMD as strategically narrow in 1976; only later attempting to make it explicit that, even if x86 had turned out not to be temporary, AMD was still supposed to be.
it's interesting that it points out that the agreement predates the 8088 which was the first intel CPU to even contain microcode, so perhaps it wasn't even on their radar.
AMD didn't have to copy the 8088 and 80286 as they were second-source manufacturers and so Intel just gave them the die masks, everything soured when AMD wanted to make a 287 which for whatever reason Intel hadn't hired them to produce
@fuzzyfuzzyfungus the whole second-source CPU thing was, from what I understand, an external requirement of IBM, who wanted to make sure there was not going to be supply issues getting CPUs for their PC products.