The iAPX 432 failed for a very good reason: its performance was infamously abysmal. The performance of the iAPX 432 is actually the subject of one of my favorite systems paper of all time: Robert Colwell's "Performance Effects of Architectural Complexity in the Intel 432"[1] -- a paper that I love so much that I wrote a reasonably detailed review of it decades after it was published.[2]
I love studying failure, or at least when systems break. Please do a follow up that addresses Itanium and the P4.
The failures you describe in the post were technical manifestations of an operational failure in bringing an engineering project to completion. The scope was too large, with too many unknowns. Had they scheduled their project relative to Moore's Law, they could have kept the team smaller until they had the transistor budget to ship the chip they designed. You should read about how the Aztek was made [1]
From my point of view I guess it actually seems like a terrible reason in the long run, given that even something as widely derided as the x86 architecture has, over time, been made to perform.
The bicycle is an excellent technology that has largely failed in western world, but not because it is a bad idea. The car economy has more money and more political power, allowing it to displace the bicycle.
I wasn't making a value judgement on bicycles. Look at the growth curve and adoption rate of bicycles in the western world. The bicycle paved the way for the car. Tubed tires, chains, sealed roads. All for the bicycle. Cities could have been denser, cleaner, quieter and vastly safer with bicycles. Places like Amsterdam weren't always cycling paradise, canals were getting paved for to make streets for cars. In China which has had great bicycle adoption is seeing a state sponsored push to switch over to a car based consumption economy.
Only recently has the bicycle seen a resurgence in the west. Because there is a small uptick doesn't mean bicycles as a technology have succeeded the level they should have compared to the alternatives.
But certainly processors can do more than they do. For example, doesn't x86 offer a memcpy or memmove instruction and just decode it as the right "fast" way, versus making people update their stdlib with complicated code?
I don't really know what I'm talking about, but it seems I often run across long discussions where people are trying all sorts of instruction sequences and they're model-specific and it just seems like there should be a few more higher-level instructions exposed. (Well perhaps that's the point of all the SSE and AVX instructions.)
The failures you describe in the post were technical manifestations of an operational failure in bringing an engineering project to completion. The scope was too large, with too many unknowns. Had they scheduled their project relative to Moore's Law, they could have kept the team smaller until they had the transistor budget to ship the chip they designed. You should read about how the Aztek was made [1]
[1] http://www.roadandtrack.com/car-culture/a6357/bob-lutz-tells...