Why is using stack data not suitable for implementing multithreading?

Stack data are local variables and the execution context of a function, stored in each thread's stack. In multithreading, each thread has its own stack, isolated from others.

Using only stack data is not suitable for inter-thread communication because data in one thread's stack is inaccessible to other threads. To exchange information between threads, data in the heap or other shared memory areas with synchronization mechanisms are needed.

Additionally, the stack is limited in size and intended for temporary data, whereas multithreading requires management of shared resources and states, which cannot be implemented using only stack data.