AMD has announced details of its upcoming Instinct MI355X AI accelerator, which is expected to be released in the second half of 2025. The chip is based on the new CDNA4 architecture and will be manufactured by TSMC using the 3-nanometer process. With up to 288GB of HBM3E memory and a performance of 2.3 petaflops at FP16 and 4.6 petaflops at FP8, it will be significantly more powerful than its predecessor, the MI300X. AMD is also introducing new data formats such as FP4 and FP6, which are expected to deliver up to 9.2 petaflops. AMD is aiming for similar performance to Nvidia's upcoming Blackwell B200. At the same time, AMD announced the official start of production of the MI325X, albeit with a reduced memory of 256 GB instead of the originally announced 288 GB.
Hub AI in practice
Artificial Intelligence is present in everyday life – from “googling” to facial recognition to vacuum cleaner robots. AI tools are becoming more and more elaborate and support people and companies more effectively in their tasks, such as generating graphics, texting or coding, or interpreting large amounts of data.
What AI tools are there, how do they work, how do they help in our everyday world – and how do they change our lives? These are the questions we address in our Content Hub Artificial Intelligence in Practice.
Ad
Ad
Ad
Ad
Ad
Ad
Ad
