SemiOT: Mourning for Classic Computing

----- Original Message ----- From: "Iggy Drougge" &lt;optimus(a)canit.se&gt; To: "Tony Duell" &lt;classiccmp(a)classiccmp.org&gt; Sent: Monday, August 20, 2001 8:11 PM Subject: Re: SemiOT: Mourning for Classic Computing <snip> The Z-80 has MORE registers, first of all because it has two complete sets ... but then it's of an earlier generation than the 6809 and 68K, which are of more or less the same generation but targeted at different markets. The 6502, BTW, though it's of an earlier generation than the 6809, is capable of significantly higher performance, again based on the technology in which it was originally designed and implemented, because, since it was so small and thrifty, internally, it could be run at much higher speeds than the much later 6809. In order to capitalize on the higher speed of the cpu, however, it did require faster RAM. This made the zero-page registers particularly interesting in the context of an extended register set, since, in a single instruction, albeit with multiple cpu cycles, it could access data stored there about as fast as any other CPU could access internal resources. It wasn't terribly unusual to see a 6502 that used memory at different rates, depending on location in the memory map. Consequently, it was possible to run programs that lived in lower memory at high speed without having a large amount of fast memory. The 6809 was targeted at more computing-intensive applications than was the prevous generation of processors, however, and, therefore, could do things more "elegantly" from a computer-science standpoint. If there had been an integrated MMU it would have been a major challenger to the 8xx86 series with its awkward segment scheme. The latter, however, did have a real 16-bit architecture, however. With its two-of-everything register set, the Z80 definitely had more registers internally than the 6809, though. Dick