Upvote:2
From your same link.
The Mark I read its instructions from a 24-channel punched paper tape. It executed the current instruction and then read in the next one. A separate tape could contain numbers for input, but the tape formats were not interchangeable. Instructions could not be executed from the storage registers.
A different tape is used for data.
Why is this?
In the Harvard architecture, code and the data the code operates on are physically separated. This prevents some catastrophic bugs and security exploits, because code can't be rewritten as data like it would be in a buffer overflow. The Harvard architecture is also more costly, because unused memory for code can't be used for extra data. Any memory that isn't used can't be used by the other section.