7 ms·
32-bit architectures have nothing to do with the size of different data types that have existed forever. We had 64-bit longs in 8-bit cpus. Also, there are per
by diego 12y ago
32-bit architectures have nothing to do with the size of different data types that have existed forever. We had 64-bit longs in 8-bit cpus.
Also, there are perfectly valid applications that require numbers of 8, 16, 32 or 64 bits (or variable encodings with arbitrary precision). Petabytes, embedded microcontrollers, etc.
- agumonkey 12y agoTrue, the HP48 pocket calculator was a "4bit" cpu with 64bit fp-able registers.
- jmount 12y agoSorry I was unclear. 32 bits architecture can mean a lot of different things (buss sizes, address word sizes, and so on). Mostly I am done with small pointers (having to use segments to address all of your memory, or not being able to memory-map a disk sucks) and small counters (only being able to put signed 32 bit integers into a collection sucks).