In the AI accelerator aspect, AMD is theoretically ahead of Apple, Intel and Qualcomm. In terms of performance, the biggest speed boost is achieved with an integrated AI block that can perform 50 trillion operations per second (50 TOPS). AMD previously advertised the Ryzen 8040 processor with up to 16 TOPS. This means the new processors are fast enough to run Microsoft Copilot natively. So laptop manufacturers can market their devices as Copilot+. Normally, companies measure TOPS numbers for AI algorithms based on simple eight-bit integer calculations (INT8). But the XDNA2 accelerator in the Ryzen AI 300 can also handle the relatively new Block Floating Point (BFP) data format for floating point calculations at 50 teraflops. Designed to combine INT8 speed with FP16 accuracy. Conventional floating point types represent each number separately with a sign (mantissa), pointer, and sign, while Block Floating Point combines multiple values and uses a common exponent for the entire block of values. AMD also saves chip space with this approach.

In artificial intelligence applications, Ryzen AI 300 is said to be faster than all other mobile processors. For example, AMD mentions Apple's M4, Qualcomm's Snapdragon X Elite and Intel Core Ultra 100 + 200. But for now, AMD is largely lacking in software support. Currently, there's also no dedicated AI interface, which means AMD will still have to fall back on Microsoft's rather poorly optimized DirectML under Windows 11. No information is available on whether the block supports Floating Point or not.