Prop II question. Undoubtedly covered before, but...
MarkS
Posts: 342
Would it be possible to design the chip so that the programmer could optionally designate an external address (I2C, maybe?) for the code (via a program counter?), so that the Prop does not load it into memory, but runs it directly from an external EEPROM? It would be nice to not have a program size limitation, even if it meant a slow down in operating speed. It would be nice to designate that Cog 0 run the code from address x to y and Cog 1 from y to z, etc.
Comments
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
E3 = Thought
http://folding.stanford.edu/·- Donating some CPU/GPU downtime just might lead to a cure for cancer! My team stats.
It might look better with SPIN: A SPIN bytecode it fetched around each µs, so we need just 8 MHz That could be accomplished with a kind of SPI.
As the PROP II will run 4 times faster, we have to consider this: it will become 3,2GHz then for COG instructions and 32MHz for SPIN to make it competative...
Once you boot-up to a minimal Spin program ( which can launch whatever it needs to ) you have far more flexibility to do whatever you want than could be provided in hardware. You could choose to run from I2C eeprom, SD Card, HDD, even download code from the internet or server to execute. And of course you can change the code as things evolve.
What you suggest could be done in software with the Propeller as it is now with a little bit of work.