AMD and Sony collectively mocked AMD’s strategy to enhancing the longer term efficiency of its graphics {hardware} in a video posted on YouTube this week: compression, aggregation and dedication. Compress all information within the graphics course of to cut back reminiscence overhead; including the compute items that course of the info for sooner matrix multiplication (key to enhancing AI efficiency, together with enlargement); and at last including devoted silicon to deal with ray tracing and trajectory acceleration, wanted to enhance visible high quality.
Sony’s involvement instantly made everybody nervous. PlayStation 6 rumor area: AMD chips energy Sony’s PlayStation consoles, and that is just about the one place the place the 2 firms intersect, at the very least for the second.
AMD powers virtually each console, from the xbox towards steam cowlwith the Nintendo Switch line one of many few exceptions (it’s based mostly on Nvidia chips). can also be in laptops and the corporate’s personal graphics playing cards. If you want gaming on a laptop computer that will not break the financial institutionBetter built-in graphics are at all times higher for you.
The three new applied sciences proven within the AMD video are:
Glow Cores: My testing over time has proven that AMD has lengthy lagged behind Nvidia on the subject of ray tracing efficiency (which is not only for fairly reflections, however considerably improves lighting), and that is at the very least partially as a result of its processing takes place within the cores of its fundamental compute unit, that are optimized to course of different varieties of graphics. Therefore, ray tracing reduces the body price rather a lot. And the structure of a core, a ray tracing unit limits the quantity of processing it may well do to enhance. The Radiance Cores deal with ray tracing acceleration individually, much like how Nvidia’s RT cores do.
Neural matrix: Matrix multiplication is the important thing algorithm for accelerating on-device AI processing (it is what Tensor cores deal with, for instance) and right this moment, upscaling is pushed by algorithms based mostly on machine studying and synthetic intelligence, similar to Nvidia’s DLSS and Intel’s XeSS. Upscaling is vital as a result of it is a crucial solution to run larger resolutions with out affecting efficiency, and in some ways it’s the heart of a set of applied sciences to enhance picture constancy and efficiency. AMD’s model is FidelityFX Super Resolution, and their subsequent era of know-how, Redstone FSR (in all probability a part of RDNA 5), you will want these fixes, in addition to Sony’s variation, PSSR.
Universal Compression: The much less compressed your information is, the extra reminiscence it should take to course of it and the slower it should transfer by way of a pipeline. Traditionally, graphics processors have been restricted to compressing solely the largest reminiscence guzzlers, beginning with textures, partially as a result of inserting it into the rendering pipeline got here at a efficiency value. But silicon is a lot sooner than it was once that it in all probability is smart to make use of it for all graphics information, which is how Universal Compression works. Even if efficiency is poor, it in all probability means much less reminiscence is required, an vital issue for 4K and better gaming, in addition to pricing.
This teaser is the primary of in all probability one million new applied sciences in AMD’s PS6 and RDNA 5, and I’d anticipate to listen to much more about it at CES in January 2026, if not sooner. I contacted AMD for extra particulars, however didn’t instantly obtain a response.
