How the Inside of Your Game Console Works
Whether you own a Nintendo Switch, a PS4 Pro, or an Xbox One X, you’ve got a platform capable of playing new, cutting-edge games — but what sets these machines apart from each other, or from mainstream gaming PCs? In many cases, less than you might think.
Traditional PCs are designed as generalist systems with the option to add workload-specific accelerators, like high-end GPUs. Because consoles fit into smaller spaces and are purpose-built for gaming, they can take advantage of a higher degree of component integration. Smartphones, game consoles, and PCs all use a type of highly integrated processor known as an SoC, or System-on-Chip. The SoC contains the actual CPU cores, the CPU L1 and L2 caches, a graphics processor, various I/O connectivity (USB ports, hard drives), and the memory controller that interfaces between the other function blocks and the main system RAM. In the old days, these functions were typically broken out into multiple chips on the motherboard; today they’re integrated into a single functional block.
For reference and example, here’s the motherboard off a launch PlayStation 3.
The Nvidia-built GPU is on the left, the Cell Broadband Engine (CPU) is on the right, and the chip above the CBE is the southbridge, where I/O connectivity was provided. The CPU’s XDR RAM is the four blocks of just to the right of the CBE. Compare that with the Xbox One X:
While machines like the PS3 eventually moved to a single SoC later in the platform’s lifespan, the Xbox One X and PS4 debuted with these technologies in place. The rationale is simple: The fewer chips on the board, the less complex the routing and the fewer components you have to pay to install. The Xbox One X’s central SoC is the large processor on the board surrounded by its memory. While we’ve focused on the Xbox One X and Switch as the most-and-least powerful consoles of their respective generations, these trends hold true for the PS4 and PS4 Pro as well.
What’s impressive is that you can see this same design philosophy on a device like the Nintendo Switch and on your own smartphone.
This image, from iFixit, shows the Switch’s SoC (red), the 4GB of RAM (orange), and its Wi-Fi and Bluetooth controllers (2x green boxes). What’s surprising isn’t that mobile devices are tightly integrated, but that we’ve seen this integration play out even in large systems. The same is absolutely true of PCs. While many PCs continue to offer large numbers of external expansion ports via PCI Express (thereby requiring certain minimum amounts of real estate), many to all of those connective ports use silicon built directly into AMD and Intel’s latest CPUs.
At a hardware level, PCs and consoles are more alike than ever. Switch runs on ARM, but the Xbox One and PS4 (and their upgrades) are all x86 processors that use a PC-derived graphics architecture. Practically speaking, the only difference between the Xbox, PS4, and PC is the operating system and the capabilities the developer has chosen to expose to end-users.
When the Xbox Series X and PlayStation 5 launch later this year, even some of these thin boundaries will collapse further. Up until now, console backward compatibility between generations was a strictly case-by-case affair. The PC was the only gaming system that offered a theoretically unbroken chain of support all the way back into the dim and dusty days of the command line. Now, that’s changing. Both the Xbox Series X and PlayStation 5 will launch with support for previous platforms. More significantly, they’ll both launch with promises that this support will continue into the future, not fade away after a few years.
The other major change coming with this next console generation is that we’ll once again have hardware with big-core CPUs. AMD’s Jaguar did a marvelous job powering the PS4 and Xbox One generation, but it’s time to replace it with something faster, and the Ryzen cores at the heart of both systems promise a dramatic uplift in CPU performance.