I think you misunderstand. The SA-1 isn't just a co-processor. It's basically an entire SoC in and of itself - running with the same privileges as the SNES's main CPU and having the ability to even interrupt the 65C816.
So your analogy doesn't hold - it's not a co-processor, it's a powerful second CPU with more horsepower than the main chip (~10MHz vs ~3.6MHz) and even features its own internal RAM (which also was faster than the RAM in the SNES), it had its own timers, DMA controllers, memory mapping, etc.
The SA-1 was basically an entire SoC in and of itself that could talk to the main console and use the console's resources. Just to put into perspective what we're talking about here. And yes, I would indeed consider that a port, rather than a "hack" to support some co-processor that adds faster maths.
Comments
I think you misunderstand. The SA-1 isn't just a co-processor. It's basically an entire SoC in and of itself - running with the same privileges as the SNES's main CPU and having the ability to even interrupt the 65C816.
So your analogy doesn't hold - it's not a co-processor, it's a powerful second CPU with more horsepower than the main chip (~10MHz vs ~3.6MHz) and even features its own internal RAM (which also was faster than the RAM in the SNES), it had its own timers, DMA controllers, memory mapping, etc.
The SA-1 was basically an entire SoC in and of itself that could talk to the main console and use the console's resources. Just to put into perspective what we're talking about here. And yes, I would indeed consider that a port, rather than a "hack" to support some co-processor that adds faster maths.