4 ms·
> it turns out that there are a lot of situations where PIO can avoid the need for an application-specific peripheral And I can only repeat: I think that's an
by ajross 9mo ago
> it turns out that there are a lot of situations where PIO can avoid the need for an application-specific peripheral
And I can only repeat: I think that's an aspirational delusion. I'm not aware of anyone shipping a PIO solution to anyone in volume. It's "useful weird" to Hackerspace nerds like us, and that leads to some epistemological skew.
Hardware needs to be boring and reliably supported (by people you can sue!) or else no one will bet a 10k unit PCB run on it. This is anything but.
- duskwuff 9mo agoI don't know if it's quite at the volume you're looking for, but ZuluSCSI [1] uses RP2350 (and, earlier, RP2040) for SCSI <-> SD. [1]: https://zuluscsi.com/ https://zuluscsi.com/
- ajross 9mo agoYeah, that's sort of the market that's picked it up: "I need to connect to old junk". And it would seem like a great fit, except that (as you point out) volumes are tiny and the benefit of a $1 chip vs. a $10 FPGA is high only at scale. I just checked, and that board is $70. Here's a significantly more capable device with an ESP32 and a iCE40 FPGA which could do the same things and more for $100: https://groupgets.com/products/ice-v-wireless https://groupgets.com/products/ice-v-wireless I have one and it works great, though not many people picked it up. It's actually available for $50 now, it looks like they're liquidating what they bought. Point being: is PIO really enabling new solutions in this market? No. At best it's providing a ~30% reduction in retail price for some oddball niche applications. Is that worth the cost of spinning the masks for the chip? IMHO, no.