Single-tile path is IDENTICAL to the working pre-P5 kernel. Multi-tile path uses FA2 online softmax with sOacc accumulator. Runtime branch on is_multi_tile = (n_kv_tiles > 1).
Single-tile path is IDENTICAL to the working pre-P5 kernel. Multi-tile path uses FA2 online softmax with sOacc accumulator. Runtime branch on is_multi_tile = (n_kv_tiles > 1).