How Bosses ‘Squeeze’ Your CPU: A Fun Look at Kernels, Threads, and Processes

At 9 am you sit at your desk and the boss assigns a PPT update. In this analogy you are a process , the PPT is your task , and your thinking power is the CPU .

1. Process: an executing task bundle

A process is a running task with its own independent memory space. Opening a WeChat window or a browser each creates a separate process.

# Python create a process
from multiprocessing import Process

def work():
    print("I'm working!")

p = Process(target=work)
p.start()

2. Thread: multiple workers in the same office

Threads live inside a single process, share the process’s resources, and each has its own execution state. The diagram shows a WeChat process with three threads handling messaging, file transfer, and update download.

Process (WeChat)
├── Thread 1: handles messaging
├── Thread 2: handles file transfer
└── Thread 3: handles download updates

3. Core: the real workers

A CPU chip may have 4, 8, 16 cores. Each core is an independent execution unit. Two forms exist:

Physical core : the actual hardware execution unit.

Logical core : created by hyper‑threading, allowing one physical core to appear as two logical ones.

Intel Core i9‑12900K has 16 cores (8 performance cores + 8 efficiency cores), analogous to a company with eight senior and eight junior staff.

4. Relationship among process, thread, and core

Process = an entire project team (its own resources)
Thread  = team members (share resources)
Core    = actual employees (execute work)

A 4‑core CPU can theoretically handle four heavy tasks simultaneously. If a program spawns 100 threads, the operating system schedules them round‑robin on the four cores, similar to four service windows serving many customers.

5. Exploitation patterns

Single‑process single‑thread : one employee does one job – stable but limited.

Single‑process multi‑thread : one employee handles calls, emails, etc. – efficiency increases.

Multi‑process : multiple employees each work independently – no interference.

Multi‑core parallel : several employees work simultaneously – maximum throughput.

# Multiprocessing example
from multiprocessing import Pool

def task(x):
    return x * x

with Pool(4) as p:  # 4 processes working in parallel
    result = p.map(task, [1, 2, 3, 4])

6. Summary

CPU : the overall computing power of the system.

Core : the number of actual workers that can execute instructions.

Process : independent project groups with separate memory.

Thread : members within a project group sharing resources.