Jules2, it is a little more complicated than that -
sgolds, I am/was fully aware that intel owned x86. And due to that fact they deserve royalty payments from any licensee. I may be an AMD fanboy but I'm not dishonest.
By the same token, can it not be said that AMD ownes the extentions to x86? Having said that, how many companies enjoyed succes building on someone elses technology?
There were lawsuits on this subject, resolved back in the early 90s. Once Intel reached the point where they no longer needed second sources to satisfy OEMs, Intel tried to stop licensing x86. Intel tried to claim that it owned the name '486' and the x86 instruction set and registers, but the courts ruled otherwise. So a lot of x86 is legally in the public domain, but underlying technology which implements x86 can be patented. Also, SSE (and SSEII, SSEIII) are not public domain because they are more recent, and Intel properly protected these extensions. (I couldn't tell you exactly what in SSE is proprietary because the instructions aren't. Must be how they are implemented.) Starting with K6, AMD designed their own underlying technology for x86, and starting with K7 AMD designed their own bus technology, so I believe the royalties are paid for SSEs mainly.
That is why the 586 chip was marketed as Pentium, by the way.
That is also one of the big reasons for Itanium - it is not only Intel's vision of the future of processors (EPIC), but it is also a fully protected architecture in which you can not copy the instruction set without also violating underlying technology which can be copyrighted. Thus Intel fully controls Itanium, but has very little control over x86 (except for the SSE extensions).
Bottom line: Instruction sets can not be copyrighted or patented, but implementation techniques used for instruction sets may be patented. So it is not true that Intel owns x86, and it is not true that AMD owns the AMD86 instructions or registers. Thus Intel and AMD have to cooperate with each other.
