PRISM: Breaking the O(n) Memory Wall in Long-Context LLM Inference via O(1) Photonic Block Selection

ArXi:2603.21576v1 Announce Type: cross Long-context LLM inference is bottlenecked not by compute but by the O(n) memory bandwidth cost of scanning the KV cache at every decode step -- a wall that no amount of arithmetic scaling can break. Recent photonic accelerators have nstrated impressive throughput for dense attention computation; however, these approaches inherit the same O(n) memory scaling as electronic attention when applied to long contexts.