[TRTLLM-12499][feat] (WIP) Add support for pipelined KVCache transfer for disaggregated serving in Python Cache Transceiver#15727

Open

athena-nv wants to merge 10 commits into

NVIDIA:mainfrom

athena-nv:trtllm-12499-pipelined-kvcache-transfer

Commits on Jun 25, 2026

[TRTLLM-11608][feat] Chunked KV cache transfer with early block release

chienchunhung
authored and
athena-nv
committed
fix: preserve mamba_state_index in chunked slices and fix VSWA shared counter

chienchunhung
authored and
athena-nv
committed
Address CodeRabbit review nitpicks

chienchunhung
authored and
athena-nv
committed
Address reviewer comments from eopXD and pcastonguay

chienchunhung
authored and
athena-nv
committed
Move chunk_block_offset into KVSlice dataclass

chienchunhung
authored and
athena-nv
committed
fix ups
athena-nv
committed
address coderabbit comments
athena-nv
committed
Revert early block release implementation
athena-nv
committed

Commits on Jun 26, 2026

[TRTLLM-12499][feat] Add support for pipelined KVCache transfer for disaggregated serving in Python Cache Transceiver
athena-nv
committed

Commits on Jun 29, 2026

Add integration accuracy test, and remove cuda_event from KVSlice
athena-nv
committed