It just occured to me. Your views?
Probably 99% of USARTs are used as UART (so don't need the USART mode).
Applications with high speed synchronous communiucation probably use I2S (so again USART mode is not missed).
Yup, I'd have to agree with Mark on this.
I would suggest SPI might be the new king with respect to high speed chip to chip interfaces. Look at Flash, Wireless chipsets - a lot support very high speed SPI.
But they are hardly asynchronous. The USART has been replaced by USB.. IMO.
@Mark - did you mean I2C? or I2S? I2S is kind of limited to the audio world. No reason it should be limited to that world, but it seems to be.
I did mean I2S but it depends on what factors are involved and SPI is of course often suitable.
I've used both SPI and I2S for proprietary high speed communication between processors (for example - including over optical connections which allow long distances) and personally I prefer I2S for it because the inbuilt frame synchronisation makes it easier to handle detection/loss of synchronisation in noisy environments or when the two processor can be powered independently (and so one suddenly stops or resets) [I never actually used I2S for audio, although that is of course where it originated]. SPI can do it using chip select framing but the HW doesn't coordinate it as I2S HW does, which puts a fiddly SW layer on top to do it, which tends to be more SPI controller specific.
USB is still not used much in industrial communication and is quite succeptible to noise in some environments. It is popular as connection to a configuration tool (preferred over UART) but RS485 etc. is still practical for the actual communication between devices. Ethernet of course has various advantages in many instances, which is my preference in the majority of cases.
Thanks @Mark Butcher and @Scott Swanson
Retrieving data ...