| HN Mirror

If we're talking about the ability to repurpose salvaged chips for new purposes, it barely matters what the instruction set is. What matters enormously about the 8051 in particular is that (1) by default it runs code from mask ROM you can't change, unlike the 8751, and (2) it has a pin called "external access" which lets you force it to run code from external SRAM or EPROM. Neither (1) nor (2) is generally true of the clones. What matters most then is whether you can figure out how to get it to run your code and toggle its pins. Can you load code via JTAG? UPDI? Changing an external ROM chip? Maybe it boots from an SPI Flash? That's what matters most.

(It also matters whether you need things like an external crystal or three power rails.)

I mean, it's convenient when you can use an existing compiler, or at least can find documentation for the instruction set; and 8-bit instruction sets and address buses are constraining enough that they can really make it hard to do things. But these are not nearly as important as being able to get your code running on the chip at all.

So, no, instruction-set-compatible clones (like the ones I mentioned I found in https://news.ycombinator.com/item?id=43488079, the day before your comment) are not interchangeable with Intel 8051s in the context of improvising computational capabilities out of garbage. Pinouts matter. Programming waveforms matter.

With respect to the Z80-clone TI calculators, in https://news.ycombinator.com/item?id=43488344 Virgil explained that they can in fact run CollapseOS, but can't yet self-host because CollapseOS can't yet access their Flash. If you want to use them to control motors or solenoids or something, you still need some pinout information, which I'm not sure we have.