Whatever the wording, there is a crucial difference between the two, irrespective of whether or how frequently they overlap or intersect.
There is a fundamental difference between asynchronous separation of concerns (concurrency) and multiple simultaneous use of resources (parallelism). Concurrency is a conceptual organization, whereas parallelism is a physical organization. These are orthogonal and you're free to have one, the other, neither, or both (in different ways). Just because they can interact in interesting ways does nothing to diminish their difference.
The point the OP was making is correct and well phrased, the comments here not withstanding. People who come from C land are used to dealing with threads (concurrency) as the primary means of doing anything non-serial. Some may be familiar with SPMD message-passing (parallelism), but they'll already know what's up. Unless you've programmed in Fortran or APL, are familiar with the Connection Machines, or do GPU coding, you won't have experienced SIMD (parallelism). People who've only ever used threads have an extremely impoverished view of the world of parallelism. Because they know threads==concurrency and they know threads can be used for parallelism, they assume the two are coextensive when they really are not.
There is a fundamental difference between asynchronous separation of concerns (concurrency) and multiple simultaneous use of resources (parallelism). Concurrency is a conceptual organization, whereas parallelism is a physical organization. These are orthogonal and you're free to have one, the other, neither, or both (in different ways). Just because they can interact in interesting ways does nothing to diminish their difference.
The point the OP was making is correct and well phrased, the comments here not withstanding. People who come from C land are used to dealing with threads (concurrency) as the primary means of doing anything non-serial. Some may be familiar with SPMD message-passing (parallelism), but they'll already know what's up. Unless you've programmed in Fortran or APL, are familiar with the Connection Machines, or do GPU coding, you won't have experienced SIMD (parallelism). People who've only ever used threads have an extremely impoverished view of the world of parallelism. Because they know threads==concurrency and they know threads can be used for parallelism, they assume the two are coextensive when they really are not.