A common analogy used for describing a computer’s processor or CPU is thinking of the computer as a human body, and the CPU is its brain.

From this point forward we’ll refer the processor as a CPU and vice versa.

The CPU is in charge of processing the computer’s tasks by constantly resolving mathematical and logical problems.

Although electronic computers have been around since World War II’s ENIAC, the ability to put their CPU (Central Processing Unit), into a single circuit board and chip became a reality only in the seventies, with the Intel 4004 being the very first commercially available microprocessor.

The Intel 4004 had a 4-bit architecture, allowing an addressable memory capacity of 640 bytes, that is, RAM capacity, and a maximum clock speed of 108 kilohertz.

The clock speed is how many times per second a CPU completes a work cycle, in this case, we are talking of 108 thousand cycles per second.

Of course, these specs aren’t that great if we think of modern CPUs.

By comparison, a cheap Intel i3-9100F can complete more than four billion cycles per second, per core.

The basic CPU processing system cycle goes like this: fetch an instruction from Random Access Memory, decode it, and execute it, over and over again.

From the 70’s to now, what changed the most in processors was speed.  This was made possible by manufacturers’ ability to introduce more and more transistors into the same space.

However, what a processor does remains the same: reading, processing, and moving information in the form of bits, in nanoseconds.

Despite its age, the Intel 4004 processor had mostly the same components as today’s CPUs:

The arithmetic logic unit or ALU is in charge of performing arithmetic operations as well as logical functions.

The control unit or CU, organizes the CPU instructions and tells the ALU, the memory, and the input and output devices what to do.

Lastly, the CPU Registers, which are a form of memory, hold results from former operations required for subsequent ones. Built inside the processor, the registers are on top of the memory hierarchy.

As mentioned earlier the Intel 4004 was a 4-bit processor. This doesn’t mean that it was unable to process larger numbers, but rather that it could only work with 4 bits at the time.

What is a bit? You might be asking.

A bit is a “binary digit”.  It is the basic unit of information, which can be either 1 or 0. In the physical world that means a bit can represent “high voltage or low voltage”.

Bits act as building blocks for larger pieces of information and are used in binary code.

Binary code is a computer language based on “yes or no” or “true or false” logic.

The use of binary enables the CPU to perform logical operations, specifically Boolean algebra, for interpreting and resolving instructions.

Most modern processors are based on the x86 Instruction Set Architecture that is featured in Intel’s 8086 chip from the late seventies.

The CPU architecture is how the CPU instruction set is designed.  It is a type of internal computer language.

Since its introduction, the x86 architecture has incorporated many more instructions.

Some of its most basic commands are LOAD, for loading information from RAM, IN, for inputting information from a device, and ADD for adding two numbers.

Over the last two decades, 64-bit architecture processors have become the norm.

64 bits are the number of bits the processor can process at the same time.

If the Intel 4004 could only work with 4 bits at a time, a 64-bit processor can deal with 64 bits at a time.

The CPU architecture impacts the number of bit locations in RAM that the processor can access. So that means 32-bit architecture was limited to 4GB of instructions, and 64-bit can theoretically manage up to 2^64 bits, that is 18 exabytes or around 18 million terabytes, an immense leap.

CPUs also have cache memory, which only became mainstream in the early nineties and today can be found divided into three levels: L1, split into L1 for storing data and L1 for storing instructions.  There is also L2 and L3. L2 is faster and lighter, and L3 is slower but has more storage capacity.

In a multi-core processor, each one of its cores has its own L1 and L2 caches. Multi-core processors are independent processors put together in the same integrated circuit each with its own instructions and programs.

Modern CPU’s also have a register memory system.  This type of CPU memory differs from CPU cache, since the registers hold a small number of instructions for immediate use, while the cache stores a greater number of repeated instructions.

Due to its capacity for managing multiple tasks, multi-core processors have become today’s standard for computers and even smartphones.

Before the introduction of multi-core processors, MultiThreading was a clever way of optimizing the CPU’s workload. So much so that this technology continues to be present in today’s CPUs.

AMD’s Simultaneous MultiThreading and Intel’s HyperThreading are both based on the MultiThreading technology introduced in the nineties.

MultiThreading allows each of the CPU cores to perform two sets of tasks or threads at the same time.

We shouldn’t look at these threads as hardware, but rather as virtual cores. Despite having additional threads, each core is still only able to do one thing at a time.

However, additional threads gives the CPU cores additional processing capacity to enable them to manage tasks in parallel.

Ok, enough theory let’s look at some practical stuff.

The motherboard holds the CPU using a socket, which provides electricity to the processor using pin connectors.

Depending on the processor and motherboard the pin connectors can be one of two types: Land Grid Array, where all pins are in the socket, or Pin Grid Array where all pins on the CPU. Damaging only one of these pins can damage the CPU or motherboard.

Most sockets on the motherboard have a locking system in which the CPU is attached by clips and closed by a latch.

This mechanism is designed for applying force constantly between the CPU and socket, while also allowing for the CPU to be replaced.

This is not the case, on many laptops and generally on cell phones, where the CPU is welded to the motherboard.

As a result of the countless cycles they complete per second, CPUs generate a great deal of heat.

Initially, processors dissipated heat through a heatsink to which they were bonded by thermal paste to optimize the heat transfer.

Now, CPUs have a different system for cooling down, starting from the design of the PC case itself, using multiple fans and water-cooling systems depending on the level of heat generated.

When understanding how a computer works, Processors are important.  The next important thing to understand are Chipsets.