Subject: Re: None
To: None <email@example.com>
From: None <firstname.lastname@example.org>
Date: 12/03/1995 23:05:52
>>>>> "Joe" == Joe Torre <email@example.com> writes:
Joe> I dont know if the older SunOS could run on *any* Amiga, because
Joe> you would need the 68010 for sure. But an A500 can easily have
Joe> 1meg ram, and the 68010 is a drop in replacement for 68000 Amiga
The Sun1's started using 68000's according to the hardware
FAQ, but not many were made.
Joe> However, The Sun 2 is a 68010 for sure. Sun Model 2/120 = 10mhz
Joe> 68010-1meg ram 2/170 = 10mhz 68010-2meg ram
Joe> The 68010 has a way of emulating virtual memory by traping
Joe> access to memory areas not physically present by using the vector
Joe> base register (VBR).
If I'm not mistaken, the VBR allows you to have a duplicate
set of interupt vectors and use them instead of the hardwired ones
(which lie in the first few pages of RAM). To the Amiga, this meant
that your interupt handing vectors could be in FAST ram instead of
chip, which made things faster.
However, even if what you say is true, you still have the
problem of REAL memory getting accessed and that what you're proposing
would cause an interupt for every single virtual memory access.
Joe> The Sun 3 has a 68020 CPU and 68451 MMU. Any Motorola 020 systems
Joe> could be upgraded to 030 by installing a hudge jumper plug in the
Joe> 68451 socket and a mezzinine board with an 68030 in the 68020
Joe> socket. This "upgrade" also removed one waitstate on any address
Joe> translations and made quite an improvement. Math coprocessors on
Joe> 020 systems were with 68881 (12-33Mhz) and 030 systems used the
Joe> 68882 which allowed concurent instruction execution with the
Joe> 030. 50mhz 68882 can be installed where a 12mhz 68881 once was
Joe> (perhaps with a 60mhz xtal clock!) and the floating point
Joe> performance is *quite* improved. Render ho!
In a similar vein, if you took a 68000 or 68010 and placed it
on a daughter board, coupled a MMU with it and plugged it into the
68000 socket on an A500, it would be closer to the old suns. That
wouldn't solve, however, the other hardware differences. Besides,
when you think about it, that's what you're doing when you put an 030
or 040 board into an A500 or A2000.
|David Gilbert, PCI, Inc., Richmond Hill | Two things can only be |
|Mail:firstname.lastname@example.org | equal if and only if they |
|http://www.pci.on.ca/~dgilbert | are precisely opposite. |