2 ms·
Atari ST (68k at 8MHz) had other issues that made it sluggish. It is no Amiga. In most scenarios, most interrupts aren't going to interrupt divisions, as most
by snvzz 2y ago
Atari ST (68k at 8MHz) had other issues that made it sluggish. It is no Amiga.
In most scenarios, most interrupts aren't going to interrupt divisions, as most instructions aren't over 16 cycles[0], thus most interrupts will not be much longer than 44 cycles. In microseconds, a non-issue for a microcomputer.
Realtime programmers of the time had predictable behavior, and could simply avoid using the specially long instructions (movem and div) altogether if they needed to bound latency below the upper limit. Performance isn't as important as meeting hard deadlines.
0. https://wiki.neogeodev.org/index.php?title=68k_instructions_timings https://wiki.neogeodev.org/index.php?title=68k_instructions_...