The great thing about Amigas was they had a flat 32-bit memory architecture from day one. Unlike 16-bit DOS PCs, memory was memory–to an extent. Amigas did have three types of memory. So let’s look at Amiga chip RAM vs fast RAM vs slow RAM.
Amiga chip RAM was visible to the CPU as well as the sound and video chips. Fast RAM and slow RAM were not, but fast RAM, which sat higher in the CPU’s address space, could run programs faster than either slow RAM or chip RAM.
What was Amiga chip RAM?
Chip RAM was called chip RAM because the Amiga’s custom sound and video chips could access it, along with the CPU. Memory was expensive, so rather than dedicate tons and tons of memory exclusively to the sound or video chip, the Amiga’s designers opted for versatility. They set up a pool of memory that the CPU and the custom chips could access. This saved a lot of copying data around, and it increased the system’s versatility. The video chip could use up to 512K of RAM, at a time when 512K was more memory than some entire computers had. But if you weren’t running graphics-intensive software, you could use that memory for anything else.
The result was a computer with more versatility than anything that came before it, at least at a price point consumers could afford.
The only drawback to chip RAM was that the CPU had to share the bus with the other chips. So you incurred a slight speed penalty when you used it as general purpose RAM. But that was much better than running out of memory. The versatility was more than worth it.
Early Amigas could use up to 512 KB of chip memory. The Enhanced Chip Set (ECS) increased that to 1 MB, and the Amiga 3000 and AGA-based Amigas could use 2 MB 0f chip RAM.
What was Amiga fast RAM?
Fast RAM was dedicated to the CPU. It sat higher in the 68000’s address space, starting at 0x200000, and adding fast memory to an Amiga that didn’t have any really did make the computer noticeably faster. Not twice as fast, mind you, but you noticed the difference. By putting the command FastMemFirst in your Amiga’s s:startup-sequence file or in your startup drawer, you could make an Amiga allocate RAM to programs out of fast memory and save slower chip memory for last.
What was Amiga slow RAM?
I don’t remember seeing the term slow RAM back in the Amiga’s heyday, but maybe people used it. Slow RAM refers to memory that an expansion board placed in the no-man’s land between your chip RAM and 0x200000. It was slow because it sat on the shared bus, but less versatile since the custom chips couldn’t use it. The A501 expansion board for the Amiga 500 when used in an Amiga 500 with the original chipset is the most common example of slow RAM.
Slow RAM isn’t useless, of course. If you upgraded to a newer memory controller (Agnus) that could handle more RAM, you needed that memory sitting in that address space so the custom chips could use it.
Later Amigas that could use a full 2 MB of chip RAM didn’t have any slow RAM. Amiga 1000 and early Amiga 500 models were the machines most likely to end up with slow RAM. Memory connected to the side slot of an A500 or A1000, or a Zorro slot in a big-box Amiga, worked as fast RAM.
Amiga chip RAM vs fast RAM vs slow RAM: in conclusion
Amiga memory was less confusing than DOS memory, certainly. So hopefully that clears up a common point of confusion for Amiga newcomers. If it helps, here’s a primer on Amiga DOS commands as well, and the Amiga LHA file compressor. The Amiga is very different from other computers, which makes it confusing, but also makes it interesting.