Why we moved from 32-bit to 64-bit processors?

When I had just entered college, I got my first laptop almost ten years back. Surprisingly, it said 64-bit instead of 32-bit, which I had seen till then on the computers everywhere. Why did it say 64-bit? Almost all the computers gradually moved from 32-bit to 64-bit processors in the late 2000s and early 2010s. What exactly did change with this?

Let's start with how processors exactly work. Processors are a collection of billions of transistors placed on a silicon wafer. These transistors combine to build logic gates and even sophisticated computational logic. There are two main parts to CPU- control unit (CU); and arithmetic and logical unit (ALU). CU is like an orchestrator at the opera (orchestrating the execution of programs), and ALU is like vocalists and instrumentalists at the opera (carries out arithmetic and logical computations). To carry out any computation, the processor needs some data.

We already know there are two kinds of memories in our computers- primary memory (also called RAM or just 'memory') and secondary memory (also called storage). RAM holds data and programs that processors require for executing active programs, whereas the storage persists everything else that is not being actively used. Now, RAM has the entire Microsoft Excel program and required data, but all of it is not being processed simultaneously; the instructions currently being processed are stored in registers, a tiny memory unit embedded within the processor. Technically, registers are the smallest data holding elements.

When a computer or processor is marked as 32-bit or 64-bit, it is being talked about the size of registers in the processor. There are several kinds of registers; not going deeper into these, but some hold the address of data in memory, some hold instructions, and so on. In RAM, memory is typically marked by addresses from 0 - X (a 4GB ram has 2^32 bytes and requires 32 bits to store the address on a location, whereas an 8GB ram has 2^33 bytes so on).

A 32-bit register won't ever be able to access any location in RAM beyond 4GB because it cannot store its address. This is a constraint if we want to use 8 or 16GB RAM. This is primarily why manufacturers decided to move to a 64-bit register that adds support for the address of 64-bit, i.e., allows 2^64 bytes in RAM (~16-17 exabytes; FYI, 1 exabyte is roughly 1 billion GBs). Today, with a 64-bit register, we are using at max 42-48 bits of these 64 in computers.

Along with allowing more memory, this also allowed us to have 64-bit instructions. Since instructions are a part of the operating system, they also need to support 64-bit instructions, so almost all operating systems were rebuilt for a new generation of processors. You can run both 32-bit and 64-bit OS on a 64-bit processor but only run 32-bit OS on a 32-bit processor.

That's all for today. I hope this was a tiny bit helpful.