Well, that would be interesting, as the A3000 AFAIK requires MMU for the Zorro III bus to work.
The MMU fixes a cache bug in the 68030 (with longword writes as I recall). The (data?) caches can be turned off instead but that reduces performance more than using the MMU. It's not just the Apollo core which can have bugs
.
The biggest concern is the %&$#?@!%&$#?@!%&$#?@!8364;100 price hike from the 40 to 50mhz. If 030s are really getting that expensive then maybe it time for Jens to think about FPGAs too or get left behind.
Jens has FPGA knowledge from the Chameleon and Clone-A projects. FPGA code no doubt requires more development and support time than an already developed hard CPU. He could be waiting for FPGA cores like the TG68 and Apollo to mature and the source to become available but then he is taking a gamble that current projects like Majsta's Vampire won't undercut him as bugs are fixed. The FPGA CPU cores are not only much faster but they can be *more* compatible with older 68k code than a 68040/68060. Who would pay this much for a 68030 when an FPGA accelerator for half the price is several times faster, has an FPU and has the same compatibility as a 68020? Jens could be put out of the Amiga accelerator business with a few more 68k FPGA core bug fixes. He also could be put out of the Amiga motherboard business if anyone wanted to make a standalone FPGA Amiga board.
Yes it very expensive for a faster 030.
A fast 030 without FPU no less.
