Delete comment from: Ken Shirriff's blog
Interesting you mention the lack of microcode for Zilog Z8000.
The lack of microcode (I think 18k transistors, probably the lowest transistor count of all the 16bit micros) made a radiation hardening a lot easier than others and the chip found a niche in military systems due to it's ability to do 32 bit and 64 bit maths and on a clock for clock basis competitive performance against the 68K.
Apparently Shinta who implemented the design had a lot of trouble fighting bugs due the complicated decoder and there is an IEEE article he wrote about it that I know about but is pay walled and have never read:
https://www.sciencedirect.com/science/article/abs/pii/0142694X81900065?utm_source=pocket_mylist
According to the computer history interview on the Z8k design (lost the link) the initial plan for was an introduction in 1978 at the same time as the 8086 but was delayed to 1979 just before the 68k. Searching usenet people were still complaining about the first issues with the chip but also about Zilog being a small company and not getting support.
The Zilog Z8000 deserves more attention. It was a direct competitor with the 8086 and shared the multiplexed address/data bus and had an upgrade path from the its 8bit forebear via assembly source code translators. It was squeezed out between the 8086 when IBM chose it for the 5150 PC and 68000 which was adopted for high end Unix systems.
https://www.eejournal.com/article/in-memoriam-dr-bernard-peuto-architect-of-zilogs-z8000-and-z8/
Dec 4, 2022, 9:42:31 PM