Kernel threads are supported and managed directly by the OS

Most modern OSs doesn’t support kernel threads (e.g., Windows, Linux)

Threads cannot run in parallel on multi-core systems

Multiple threads can run in parallel on multi-core systems

Multiple threads can access the kernel at a given time

Does not suffer from the drawback of the many-to-one model

Does not suffer from the drawback of the one-to-one model

Increasing the number of user threads does not affect system performance

Most operating systems use the

In practice, it is difficult to implement the many-to-many model

No need to limit the number of kernel threads on modern multicore systems

Modern concurrency libraries use the

Developers identify tasks and the library maps them to kernel threads

When a thread makes a blocking system call, another thread can run

OS creates a set of kernel threads (less than or equal to user threads)

When a thread makes a blocking system call, another thread can run

OS creates a kernel thread per user thread (can burden performance)

The entire process will block if a thread makes a blocking system call

Thread management is done in user space by a thread library not OS

Only one user thread can access the kernel at a time

The server creates a thread for each request