86 Commits
Author SHA1 Message Date
Zhiqiang Xie 7a6c652c77 [HiCache] Auto-size the host pool to fit available host memory (#40135) 2026-09-19 12:50:43 -07:00
Zhiqiang Xieandcctry ceb1d2e580 [PD] Enable optimistic prefill with buffer-only L3 write-through HiCache (#40043)
Co-authored-by: cctry <csycfl@gmail.com>
2026-09-18 16:04:15 -07:00
Zhiqiang Xie a0534f8cca [HiCache] Stop arming a prefetch retry for a too-short storage span (#40042) 2026-09-18 15:33:58 -07:00
Zhiqiang Xie 4e9e407d37 [HiCache] Label radix-cache metrics per rank and split the "shrunk" prefetch reason (#39280) 2026-09-15 17:54:04 -07:00
Zhiqiang Xie 49f540f3c0 [HiCache] Optimize buffer-mode storage existence bookkeeping (#39480) 2026-09-15 10:49:58 -07:00
Zhiqiang Xie 7f5dd19256 [HiCache] Rework the buffer-mode storage prefetch pipeline and retry bookkeeping (#39283) 2026-09-15 10:48:49 -07:00
Zhiqiang Xie beaf3d9252 [HiCache] Replace skip_lock_node_ids with a segment lock protocol (#36848) 2026-09-09 14:55:51 -07:00
Zhiqiang Xie b44496389c [HiCache] Count hit allocations and in-flight backups in the buffer pipeline idle check (#37883) 2026-09-03 16:08:50 -07:00
Zhiqiang Xie a480f388b2 [HiCache] L3 storage prefetch lifecycle metrics and cross-tier attribution fixes (#37503) 2026-09-03 16:00:04 -07:00
Zhiqiang Xie d0c95f6c91 [HiCache] buffer mode: anchor-lock staged prefetches by default (#37464) 2026-09-03 12:08:22 -07:00
Zhiqiang Xie 68978f8d52 [Scheduler] Count the parked chunked-prefill request in the busy mem check (#37502) 2026-09-03 12:08:17 -07:00
Zhiqiang Xie a328c19c81 [HiCache] buffer mode: decide staged-fetch fate against the live tree (#36834) 2026-08-29 01:51:47 -07:00
Zhiqiang Xie f611e0c073 [HiCache] Fence load-back behind the forward stream (#36738) 2026-08-28 11:12:51 -07:00
Zhiqiang Xie c9bba091f8 [HiCache] Key storage prefetch by the request namespace (#36382) 2026-08-28 11:12:10 -07:00
Zhiqiang Xie 3785b2d20f [HiCache] Reject load-back specs that claim nodes pinned by an in-flight load-back (#35931) 2026-08-28 01:59:06 -07:00
Zhiqiang Xie 1061e34785 [HiCache] Retry L3 storage prefetch after a missed attempt (#36227) 2026-08-28 01:35:21 -07:00
Zhiqiang Xie 3f1031d697 [HiCache] Heal the storage existence cache on a hybrid prefetch discard (#36386) 2026-08-28 01:33:40 -07:00
Zhiqiang Xie c35683fda0 [HiCache] Clamp tombstoned SWA locs in UnifiedSWAKVPool translation (#35933) 2026-08-21 23:08:56 -07:00
Zhiqiang Xie 0db2bdfec5 Fix buffer-mode HiCache load-back ownership races; add optional prefetch anchor lock (#35769)
Signed-off-by: Zhiqiang Xie <zqx@meta.com>
2026-08-21 20:20:36 -07:00
Zhiqiang Xie 01814e110d [HiCache] Simple style change for buffer mode (#35574) 2026-08-19 15:23:21 -07:00
Zhiqiang Xie 977412ae61 [HiCache] Buffer-only mode for HiCache host memory layer (#34798) 2026-08-18 19:21:24 -07:00
Zhiqiang XieandTingwei Huang 5469faec45 HiSparse: shared-index (IndexShare) plan-then-IO swap-in prefetch (#34329)
Co-authored-by: Tingwei Huang <huangtingwei9988@gmail.com>
2026-08-11 01:58:28 -07:00
Zhiqiang Xie 3dc91366ac [HiCache] write_back: reclaim duplicated host copy first under host pressure (#33777) 2026-08-07 15:59:59 -07:00
Zhiqiang Xie f6de147b8d Remove revoke queue after hit-then-alloc refactoring (#33613) 2026-08-06 03:10:10 -07:00
Zhiqiang Xie 106bcc1293 Observability enhancement for HiCache (#32388) 2026-08-05 14:13:06 -07:00
Zhiqiang Xie 5da0b6ec39 Write-back policy fix for unified tree (#31845) 2026-07-24 09:58:18 -07:00
Zhiqiang Xie f19246e59a [HiCache] write_back policy refinement (#29817) 2026-07-02 12:16:03 -07:00
da0f4f6f92 [HiCache] remove large host mem constraint (#28614)
Co-authored-by: Teng Ma <stmatengss@gmail.com>
Co-authored-by: Vladislav Nosivskoy <vladnosiv@gmail.com>
2026-06-26 16:20:48 -07:00
41c7c97ff3 fix hisparse LRU policy (#22170)
Co-authored-by: huangtingwei9988 <huangtingwei9988@users.noreply.github.com>
Co-authored-by: hzh0425 <hzh0425@users.noreply.github.com>
2026-04-05 18:47:58 -07:00
30ba1f78b0 Hisparse Minor Fix (#22131)
Co-authored-by: huangtingwei9988 <141888744+huangtingwei9988@users.noreply.github.com>
Co-authored-by: hzh0425 <58988019+hzh0425@users.noreply.github.com>
2026-04-05 16:15:47 -07:00
Zhiqiang Xie 9eb75211b1 style refinement for hisparse (#21198) 2026-04-01 01:03:17 -07:00
Zhiqiang Xie 13f4f010d8 HiSparse for Sparse Attention (#20343) 2026-03-22 23:09:31 -07:00
Zhiqiang Xie d9ed80b9f1 fix AMD CI failure of NUMA binding (#17184) 2026-01-16 00:21:04 -08:00
Zhiqiang Xieandhxie 675acecec6 Attention backend selection bug fix for hicache (#16779)
Co-authored-by: hxie <haiyong.xie@ustc.edu>
2026-01-10 00:22:41 -08:00
Zhiqiang XieandHao Chen cfcf2758aa [HiCache] Critical fix to host memory double free (#13501)
Co-authored-by: Hao Chen <cighao@gmail.com>
2025-11-18 01:34:23 -08:00
Zhiqiang Xie 3d40794fcf [HiCache] Cleaning the deprecated host memory state (#10778) 2025-09-25 14:43:53 +08:00
Zhiqiang Xie 3b99f23c44 [Bugfix] Retract not releasing enough memory when page size > 1 (#9989) 2025-09-07 21:41:50 -07:00
Zhiqiang Xie 0b8c5721f1 [HiStorage] Remove delete and clear as necessary methods (#10039) 2025-09-06 10:27:26 +08:00
Zhiqiang Xie 369b143366 [HiCache] Minor fix on file storage backend (#9869) 2025-09-02 15:52:37 -07:00
Zhiqiang Xieandpansicheng 8b6966d020 [HiCache] Storage Refactoring (#9797)
Co-authored-by: pansicheng <27603155+pansicheng@users.noreply.github.com>
2025-08-31 22:58:21 +08:00
Zhiqiang Xie f9076a5a2c hot fix for mooncake batch set api (#9836) 2025-08-30 21:01:51 -07:00
Zhiqiang Xie 54e872d343 [HiCache] resolve conflict between chunked-prefill and hicache hit count (#9776) 2025-08-30 01:30:54 +08:00
Zhiqiang Xie 001f51940a [HiCache] change the default policy to write through (#9772) 2025-08-28 18:28:39 -07:00
Zhiqiang Xie 43de1d7304 HiCache Storage fix host memory leak (#9648) 2025-08-26 10:49:40 -07:00
Zhiqiang Xieandgemini-code-assist[bot] 0eec4cb6cc HiCache, add bench long context plus minor fixs (#9086)
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
2025-08-11 16:54:52 -07:00
Zhiqiang Xie 9f78f391ae HiCache Storage: generate hash when inserting new nodes (#9053) 2025-08-11 14:18:59 -07:00
Zhiqiang Xie b32792516a REVIEWERS.md typo fix (#9048) 2025-08-10 22:33:37 -07:00
Zhiqiang Xie 6e0b646832 HiCache Storage tp fix (#8878) 2025-08-09 01:16:51 -07:00
Zhiqiang Xie b0add2da00 HiCache storage, style change and bug fix (#8719) 2025-08-03 15:05:04 +08:00
Zhiqiang Xie dd7ca00601 Interface change for kvcache io to support page first layout (#8318) 2025-08-01 11:37:49 +08:00
Zhiqiang Xie 9305ea6c2d HiCache, fixing hash value indexing (#8636) 2025-08-01 11:29:51 +08:00
Zhiqiang Xie 528bd1ed85 HiCache, check before terminate prefetching (#8372) 2025-07-26 23:13:16 -07:00
Zhiqiang Xie d40846d456 breakdown kernel update (#8334) 2025-07-25 08:33:17 +08:00
Zhiqiang Xieandpansicheng 145482f422 HiCache Storage TP Refinement (#8307)
Co-authored-by: pansicheng <sicheng.pan.chn@gmail.com>
2025-07-25 08:31:47 +08:00
Zhiqiang Xieandpansicheng f39037fffb HiCache Fix (#8288)
Co-authored-by: pansicheng <sicheng.pan.chn@gmail.com>
2025-07-23 16:51:32 +08:00
Zhiqiang Xie ce86e201df bug fix and tag (#8282) 2025-07-23 16:50:31 +08:00
Zhiqiang Xie b43263307f Hicache IO kernel refactoring (#8264) 2025-07-23 16:49:03 +08:00
Zhiqiang Xie 9d33fcfb8e Hicache Storage Layer Prototype (#7704) 2025-07-18 15:20:19 +08:00
Zhiqiang Xie 2fc824b84c Kernels for efficient KV cache IO (#7313) 2025-07-06 22:53:36 -07:00
Zhiqiang Xie f9eb04ddb2 upgrade sgl kernel to 0.2.1 for main (#7676) 2025-07-01 00:00:13 -07:00
Zhiqiang Xie 34c3f9b2d3 kvcache io kernels and test case (#7382) 2025-06-23 11:58:59 -07:00
Zhiqiang Xie e56685ac1b Upstreaming hicache bug fixes (#7267) 2025-06-17 17:44:57 -07:00
Zhiqiang Xie b26cb1c55a Fix problem of large page size with chunked prefill (#6046) 2025-05-06 15:19:47 +08:00
Zhiqiang Xie f8e460930a Fix prefill OOM error in the case of large page size (#5081) 2025-05-05 16:02:55 -07:00
Zhiqiang Xie 70645f4d7d upstream hicache fixes (#5570) 2025-04-20 23:08:30 -07:00
Zhiqiang Xie e2574ee986 fix hicache write back (#5543) 2025-04-19 21:56:22 -07:00
Zhiqiang Xie 9d0b36c47a fix deepgemm as well (#5030) 2025-04-03 02:41:37 -07:00
Zhiqiang Xie a2aea59b6e update cutlass tag (#5011) 2025-04-02 18:30:30 -07:00
Zhiqiang Xie 3fadc64793 bug fix for hicache host eviction (#4989) 2025-04-02 00:33:50 -07:00
Zhiqiang Xie e119f04215 Large page size aligned hierarchical caching (#4581) 2025-04-01 22:38:15 -07:00
Zhiqiang Xie 12047f5e94 Prevent memory leak of retract_decode when page_size > 1 (#4977) 2025-04-01 15:30:45 -07:00
Zhiqiang XieandLianmin Zheng a169b9f813 Fix oom error for large page size (#4913)
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
2025-03-30 21:34:21 -07:00
Zhiqiang Xie 4d25305700 Move mem_state update into debug mode (#4525) 2025-03-23 00:52:27 -07:00
Zhiqiang Xie ecbfe58bb0 Bug fix for metrics counter (#4660) 2025-03-22 13:39:21 -07:00
Zhiqiang Xie a98290aea3 Unit test for Hierarchical Caching (#4486) 2025-03-17 17:45:00 -07:00
Zhiqiang Xie f5bbf6037d Fix: Complete int32 to int64 conversion (#4465) 2025-03-16 18:14:27 -07:00
Zhiqiang Xie fbdb50501f Hot fix for hicache with new page aligned radixtree (#4397) 2025-03-13 15:50:49 -07:00
Zhiqiang Xie 10b544ae9b Hierarchical Caching Refactoring and Fixing TP issue (#4082) 2025-03-12 11:22:35 -07:00
Zhiqiang Xie 9376ac361d Memory pool fix for upstream change about eagle (#4170) 2025-03-07 00:58:20 -08:00
Zhiqiang Xie aee30630d8 Add a pointer to the real KV cache pool (#4113) 2025-03-05 21:39:07 -08:00
6c7a152c5a Hierarchical Caching for SGLang (#2693)
Co-authored-by: Wenxuan Tan <wenxuan.tan@wisc.edu>
Co-authored-by: Yineng Zhang <me@zhyncs.com>
2025-02-23 21:56:30 -08:00
Zhiqiang XieandLianmin Zheng 08104b56de Sanity check to prevent performance regression (#3171)
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
2025-01-27 12:28:17 -08:00
Zhiqiang Xie 8af7048dcf Query remaining memory dynamically for PrefillAdder (#2941) 2025-01-17 20:20:26 -08:00
Zhiqiang Xie 13387e6b7a Multi-turn benchmark for hierarchical caching (#2942) 2025-01-17 16:17:24 -08:00
Zhiqiang Xie 5d6e9467d4 Cache controller for hierarchical caching (#2804) 2025-01-10 20:22:01 -08:00
Zhiqiang Xie 51caee740f Host memory pool for hierarchical caching (#2771) 2025-01-07 21:38:37 +00:00