Talk:Motorola 68008
![]() | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||
|
I reversed an edit I noticed by a non-logged in user in this article he deleted all the info. TheYapps 03:17, 20 Feb 2004 (UTC)
I don't see a reason for the Hong Kong tag. Computer Guru (talk) 19:59, 28 February 2009 (UTC)
68008 faster than competing 8-bit microprocessors
edit"Because of its smaller data bus, it was only about half as fast as a 68000 of the same clock speed. However, it was still faster than competing 8-bit microprocessors, because the 68008's internal architecture was based around a 32-bit architecture."
It's a simple presumption that a 32-bit chip would be faster than an 8-bit chip. But that's not the experience of it. In particular, the Acorn team found this in the early 80's when looking for a successor to the 65C02. The 68008 takes 8 clocks (4 clocks per byte) to read a 16-bit instruction with no parameters. They found that their real software routines could be done faster on a 2 MHz 65C02 than on (and I'm shaky here!) an 8-odd MHz 68000. Add the fact that the 68000 interrupt response time can be very slow (current instruction must be completed first and DIVS takes 158 clocks) and you can see why they ended up developing the ARM :-)
Naturally the 68008 can deal with 16x the memory but a big percentage of applications didn't need more than 64 KB in 1982. So I think this should change to...
"Because of its smaller data bus, it was only about half as fast as a 68000 of the same clock speed. However, it still competed well with other 8-bit microprocessors and had the benefits of its 32-bit programming architecture." — Preceding unsigned comment added by TonyM101 (talk • contribs) 15:04, 28 January 2014 (UTC)
The argument seams reasonable. On the other hand one could argue that
- e.g. clearing a call stack with ADDQ.L #6, SP is much faster than POP HL; POP HL; POP HL or 6 x INC SP. Just to name one of many cases where 68k has more efficient instructions than e.g. Z80
- Naming DIVS: On 8-bit microprocessors we don't have multiplication and division, giving the 68008 an advantage
- The 68008 has more and wider registers than Z80 and 6502. This could reduce memory access significantly
My conclusion is: Without benchmarks this question can not be answered. Maybe even with benchmarks it could not be answered, I don't know. So why not leave this question open? Klabauterfisch (talk) 08:38, 18 May 2024 (UTC)