WOW, I am speechless, this is absolutely amazing. I had no idea about this project until seeing your video. I have been dreaming of a modern SE accelerator for many years, and it’s finally happening! Once the bugs are worked out and they are available, I will definitely be buying one, maybe two! Super excited for this! So much better than trying to find an old, expensive and rare accelerator that may up and die with no hope of ever repairing it.
Thanks for the video. Amiga nerd here so a brother in M68k… and I have been considering obtaining an “old world” or PPC Mac to play with. I looked up this accelerator… why did they stick with the 68000 and only 25MHz and not go with a 68040 or even better a 68060 at 50MHz (or better?)? I also understand it only has 4MB RAM, but the 68000 can use a max of 16MB (2GB for the 68040). With the 68040 you also get FPU and independent MMU. It just seems this isn’t much of an accelerator especially when compared to the Terrible Fire and especially compared to the PiStorm.
Because going that far it'd be pointless. These are designed for machines that have no video acceleration, a single 1-bit 512x342 display. There's no capability to 'add' colour on the SE. Nothing really uses the FPU in these machines - by the time software was coming out that used the FPU, they were also apps that required colour. There is no 68060 support in the Macintosh ROM anywhere - you'd have to totally re-write them for minimal performance gains. The BBU chip on the SE logic board only allows for 4MB of addressable RAM as it's just an integration of the previous PAL16 logic from the Macintosh Plus from 1986, which differed very little from the original 1984 Macintosh. The 68000 may have 16MB addressable space, but when this was released in 1987, having 4MB of ram was considered nuts and would have cost almost half of the entire machine. The rest of the address space is mapped to ROM, system overlay, expansion slots etc. Also - cost. A 68HC00020 costs a lot less than an 030 or an 040. When you butt up against the rest of the system restrictions, putting in anything more than an 020 is throwing money away.
I mean, we had not done an accelerator before so we figured we should start off simply. Our goal was a 3.5x speedup and we broke 4x in some tests so we're happy. It makes the SE much more tolerable to use. Our target price for this accelerator is $150 and in the future after the first batch of 250ish are made and sold we would like to update the design and reduce cost even more. 68030 and FPU would take up a significant part of the budget and 68060 would be totally out of the price range. I think the full '060 with FPU goes for hundreds of dollars on eBay. Even 68040s are kinda rare, but we were able to acquire about 250 of these MC68HC000s with good provenance (i.e. no fake "remanufactured" chips) for a good price. Next we will do an SE/30 accelerator which will obviously require an '030 but first things first.
@@bakkus82 They aren't, at all, but the number one priority of the Amiga nerd is being able to claim superiority because they have the biggest and baddest accelerator (bonus points if it basically just replaces the entire system through the CPU). Dude, you're still using a 68000? That's poverty spec, you should be digging through Motorola VME scrap in hopes of finding 060s for less than $500 a chip. Lame! CPU type and speed is everything!
