“Discussion of chunking sequences into non-contiguous blocks and sharing cache across requests.”