How does Tomcat interact with processor threads (hardware threads)? Why is it possible to handle 200 requests on 16 cores?

Tomcat uses a multithreading model where each incoming HTTP request is handled by a separate thread from a thread pool. These threads are software-managed by the JVM and do not directly correspond to hardware threads of the processor.

A processor with 16 cores can have, for example, 32 hardware threads (considering Hyper-Threading), but this does not limit the number of software threads that can be executed simultaneously. The JVM and the OS schedule the execution of hundreds or thousands of threads, switching them on available cores over time.

Why can 200 requests be processed on 16 cores:

• Most requests are I/O operations (reading from network, database), during which the thread is blocked and does not use CPU.
• While one thread waits for I/O, another thread can use the CPU.
• Thus, the number of requests processed simultaneously can significantly exceed the number of cores.

Example: if requests mainly wait for responses from the database, the CPU is not idle but switches between threads, providing high concurrency without needing as many cores as requests.