Persistent State Machine: Breaking the von Neumann Memory Wall for LLM Attention

A Zenodo record titled “Persistent State Machine: Breaking the von Neumann Memory Wall for LLM Attention” proposes a persistent-state-machine approach to the memory-wall problem in LLM attention. The available submission metadata points to a Hacker News discussion with a score of 2 and no comments, but does not provide authors, the paper text, hardware specifications, benchmarks, or experimental conclusions. The proposed contribution is therefore potentially relevant to AI systems and accelerator research, but its claims cannot yet be independently assessed from the supplied source alone.