They are
slower, not faster, but are extremely low power.
They are slower because each datatransfer needs handshaking
CAD tools to assist in this kind of design do exists.
I heard about the ARM996 about a year ago, but I wasn't aware that
anyone was actually selling it commercially yet. Do you know of any
implementations that are actually orderable now?
If you mean standalone, readmade CPU's, then I do not know any avaliable
commercially. The IP can be bought today.
Devices using this techniques are available : for instance the p87cl888
contains an async 8051.
People in our site (NXP semiconductors ) have used Tangram. Maybe I
should get me some prototypes.
Jos