The short version
A gaming computer is a group of specialised parts passing work to one another many times every second. Some parts store data, some decide what should happen next, some draw the picture, and some keep everything powered and cool. No single part “runs the game”. The experience you get depends on how well they hand work along.
The easiest way to understand the computer internals is to follow one frame, a single image on your screen, from the moment the game decides what it should show to the moment you see it. That path is the same in outline on most gaming PCs, even though the details vary a great deal.
Before the first frame: storage
A game begins life on a storage drive, as a large collection of files: code, models, textures, sounds and level data. When you start the game, the necessary files are copied from the drive into memory, because the drive is far too slow to feed the processor directly during play. This copying is much of what you wait for on a loading screen.
The drive does not stop working once the game is running. Many games keep streaming new pieces of the world from storage as you move, which is one reason storage speed can still matter in the middle of play.
System memory: the workspace
The data the game is actively using sits in system memory, usually called RAM. It is a fast, temporary workspace. The processor can read from it far more quickly than from a drive, but it empties when the computer powers off. Think of the drive as a filing cabinet and memory as the desk where the current work is spread out.
The processor decides what happens
Each frame starts with the processor, or CPU. It runs the game’s logic: reading your keyboard, mouse or controller, moving characters, checking collisions, running enemy behaviour and updating the state of the world. A great deal of this work is organised by the game engines that developers build on, which coordinate the many small jobs that make up a frame.
The processor then prepares a list of instructions for the graphics card: roughly, what is in the scene, where it is, and how it should be drawn. This is sometimes called issuing draw calls. How fast the processor can finish its part, and prepare that list, is one limit on how many frames per second the whole system can deliver. The details of what runs on the processor are covered in the article on processor cores.
The graphics card draws the picture
The graphics card, or GPU, takes those instructions and turns them into pixels. A GPU is built to do a huge number of similar calculations at the same time, which suits the job of working out the colour of millions of pixels. Typically it handles geometry (the shapes in the scene), textures (the images wrapped on those shapes), lighting and a range of effects.
A graphics card has its own dedicated memory, separate from system memory. The textures and other data it needs are copied there so the GPU can reach them quickly, which is why video memory matters. The GPU is also where features such as ray tracing and image upscaling are carried out on supported hardware.
The finished frame reaches the screen
The GPU writes the finished image into a buffer, a block of memory holding the completed frame. It then hands that image to the display. The display refreshes at its own steady pace, and it picks up whichever frame is ready. If the game produces frames faster or slower than the display refreshes, or out of step with it, you can see effects such as screen tearing. The relationship between what the game produces and what the monitor can show is a separate topic, covered in the article on refresh rate versus frame rate.
All of this happens again for the next frame, and the next. At a typical target of many frames each second, each frame has only a few thousandths of a second of work available for it, so every stage has to be quick.
The parts that keep it all running
Two supporting parts are easy to overlook.
The motherboard is the circuit board that connects everything. It carries the data paths between the processor, memory, graphics card and storage, and it distributes power to them. Its design decides which parts can be fitted together.
The power supply takes mains electricity and converts it into the lower voltages the components use. A weak or failing one can cause crashes and restarts that look like software faults. Cooling, meanwhile, removes the heat that all this work produces. When parts get too hot they slow themselves down to protect themselves, an effect described in the article on thermal throttling.
Why one part can hold back the rest
Because a frame passes through several stages, the slowest stage sets the pace. If the graphics card is waiting for instructions, the processor is the limit. If the processor has everything ready and is waiting for the GPU, the graphics card is the limit. People often call this a bottleneck, and the same idea appears in the discussion of hardware bottlenecks. Which part is the limit can change from game to game, and even from scene to scene within one game, since a crowded city and an empty corridor load the parts differently.
This is also why a single spec, such as the amount of memory or the number of cores, rarely tells you how well a computer will play games. Balance and the type of game both matter.
Reading a spec sheet with this in mind
With the frame’s journey in mind, a spec list becomes easier to read:
- Storage affects loading and streaming.
- System memory affects how much the game can keep ready without swapping data back and forth.
- The processor affects game logic, physics and how fast instructions reach the GPU.
- The graphics card and its memory affect image quality, resolution and smoothness.
- The display affects how much of that output you can actually see.
- The power supply and cooling affect stability over long sessions.
Treat these as general tendencies, not guarantees. Different games lean on different parts, and the only reliable way to know how a specific game behaves on specific hardware is to look at measurements for that game.