PlayStation 4 Pro GPU Hardware Customisations Revealed – Handles Two Simultaneous 16-bit Instructions

The PlayStation 4 Pro release is almost upon us. It’s also important to understand that there’s more to the console than just the TFLOPs and GBPS. Lead architect Mark wants the PS4 Pro to deliver high fidelity graphical experiences. That with double the power of the present PS4 GPU. Furthermore, it will use some new features from the AMD Polaris architecture. He also promised some tweaks to the PS4 Pro GPU. Here are the PlayStation 4 GPU details.

playstation 4 pro gpu
via redgamingtech

PlayStation 4 Pro GPU details

Sony’s PlayStation 4 Pro is capable of handling 2 16-bit floating point operations simultaneously. This is something that the original PS4 wasn’t able to do.

We know that the PS4 Pro is using AMD’s Polaris and future generations. So, the ability to handle these two 16-bit instructions will likely be very handy. Especially, when dealing with compute, and instructions which don’t require the full level of accuracy of 32-bit floats.

So, in layman’s terms, what does it all mean? Well, you’re doubling the amount of work performed. Technical precision requires the use of floating point numbers. 16 bit floating point calculations are less precise than their 32-bit counterparts and, therefore, can be performed faster. The GPU can run two 16-bit operations simultaneously, with the same amount of precision and accuracy. Now, you have the execution of a 32-bit floating point operation being handled at the same speed as a 16-bit operation. But more efficiently and accurately.

This means that a lot of operations that require on the fly calculations, such as in-game physics, can be executed even faster and better. Not all applications require this level of accuracy. So, if an operation doesn’t need that much accuracy, it will simply be executed as a 16-bit float.

Let us know what you think of this brand new reveal about the PlayStation 4 Pro GPU.