# Problem
Group the features of these two architectures.
# Process
...
# Answer
* Von Neumann.
* Single address space for instructions and data.
* Can only fetch one instruction at a time.
* Must complete each instruction before it can start on the next.
* Must alternate between reading instructions and reading or writing data.
* Harvard.
* Separate address spaces for instructions and data.
* Can fetch multiple instructions into an instruction pipeline cache.
* Can start decoding the next instruction while executing the current instruction.
* Can simultaneously read an instruction while reading or writing data.