IndexCache, a new sparse attention optimizer, delivers 1.82x faster inference on long-context AI models

IndexCache, a new sparse attention optimizer, delivers 1.82x faster inference on long-context AI models