Commits · e3d00b899f0c8a417cdfc45106c94bba5e3836a3 · OpenDAS / dynamo · GitLab

23 Apr, 2026 1 commit

feat: Tier-based KV Routing (#8380) · e3d00b89

jthomson04 authored Apr 23, 2026


Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

e3d00b89

13 Apr, 2026 1 commit
- fix(kvbm): query NCCL version at runtime instead of hardcoding (#8130) · 47c4bd46
  Alec authored Apr 13, 2026
```
Signed-off-by: Alec Flowers <aflowers@nvidia.com>
```
  47c4bd46
08 Apr, 2026 2 commits

feat: adding kvbm-engine (#6773) · 008683d6
Ryan Olson authored Apr 08, 2026
```
Signed-off-by: Ryan Olson <rolson@nvidia.com>
```
008683d6

feat(kvbm): add MLA support for DeepSeek-V2 and extend determinism tests (#7786) · ba34f131

Olga Andreeva authored Apr 08, 2026


Signed-off-by: Olga Andreeva <oandreeva@nvidia.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: wz1qqx <ziqi.wang@novita.ai>

ba34f131

06 Apr, 2026 1 commit

feat(kvbm): KVBM MLA support optimization (#7015) · 1caf6659

Aniket Kulkarni authored Apr 06, 2026


Signed-off-by: aknvda <anikkulkarni@nvidia.com>
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
Co-authored-by: jthomson04 <jwillthomson19@gmail.com>

1caf6659

04 Apr, 2026 1 commit
- fix(kv-router): migrate raw zmq paths to libzmq and refresh lockfiles (#7871) · f849e1a6
  Yan Ru Pei authored Apr 04, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  f849e1a6
02 Apr, 2026 1 commit
- feat: enable KVBM Disk allocation support different storage types (#7839) · 234a89c0
  Ziqi Fan authored Apr 02, 2026
```
Signed-off-by: Ziqi Fan <ziqif@nvidia.com>
```
  234a89c0
01 Apr, 2026 1 commit
- fix: Flaky cuda event test (#7747) · e5e65b58
  jthomson04 authored Mar 31, 2026
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
  e5e65b58
19 Mar, 2026 1 commit
- feat: make the kvbm v1's MAX_CONCURRENT_TRANSFERS and MAX_TRANSFER_BATCH_SIZE... · e7bcf651
  Richard Huo authored Mar 19, 2026
```
feat: make the kvbm v1's MAX_CONCURRENT_TRANSFERS and MAX_TRANSFER_BATCH_SIZE configurable from env var (#7527)
```
  e7bcf651
17 Mar, 2026 2 commits
- feat: enable kvbm v1 to use the v2 ::new_with_device (#7151) · 5b6624bd
  Ryan Olson authored Mar 17, 2026
```
Signed-off-by: Ryan Olson <rolson@nvidia.com>
```
  5b6624bd
- feat(kv-router): plumb medium into worker-side placements (#7462) · e811db50
  Yan Ru Pei authored Mar 17, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  e811db50
13 Mar, 2026 1 commit
- fix: harden numa cpuset lookup (#7060) · 17db1b6a
  Ryan Olson authored Mar 12, 2026
```
Signed-off-by: Ryan Olson <rolson@nvidia.com>
```
  17db1b6a
03 Mar, 2026 2 commits
- fix: pass in device_id to KVBM PinnedAllocator instead of hard code to 0 (#6809) · 0da960f1
  Ziqi Fan authored Mar 03, 2026
```
Signed-off-by: Ziqi Fan <ziqif@nvidia.com>
```
  0da960f1
- feat: upgrade rust to v1.93.0 (#6802) · 57bdbdf1
  Ryan Olson authored Mar 03, 2026
```
Signed-off-by: Ryan Olson <rolson@nvidia.com>
```
  57bdbdf1
02 Mar, 2026 1 commit
- fix: Fix vLLM KVBM integration test (#6768) · 9d10e050
  Kris Hung authored Mar 02, 2026
  
  9d10e050
24 Feb, 2026 3 commits
- feat: router metrics with dynamo_router_* {worker_id=...}. Update docs (#6227) · 1df620b4
  Keiven C authored Feb 24, 2026
```
Signed-off-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Keiven Chang <keivenchang@users.noreply.github.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
```
  1df620b4
- test: add tests for lora aware kv cache events and routing (#6523) · 6fab12be
  Biswa Panda authored Feb 24, 2026
  
  6fab12be
- feat: make KV cache events and routing LoRA-aware (#6517) · 75fea787
  Biswa Panda authored Feb 24, 2026
  
  75fea787
22 Feb, 2026 1 commit

fix: gracefully skip NIXL UCX tests when backend unavailable (#6484) · d3ee09d7

Alec authored Feb 22, 2026


Signed-off-by: alec-flowers <aflowers@nvidia.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

d3ee09d7

18 Feb, 2026 2 commits
- chore: standardizes Rust-side Dynamo environment variable names (#6358) · 56379dd5
  jh-nv authored Feb 18, 2026
  
  56379dd5
- refactor: dedup memory module with dynamo-memory; re-exports preserve current t… (#6345) · 86892a05
  Ryan Olson authored Feb 18, 2026
```
Co-authored-by: Ryan Olson <rolson@nvidia.com>
Co-authored-by: Olga Andreeva <124622579+oandreeva-nv@users.noreply.github.com>
```
  86892a05
16 Feb, 2026 1 commit
- fix: local block hash consistency in mooncake bench (#6310) · e361d6fa
  Yan Ru Pei authored Feb 16, 2026
```
Signed-off-by: PeaBrane <yanrpei@gmail.com>
```
  e361d6fa
04 Feb, 2026 1 commit
- feat: default with lib/memory, media-nixl and kvbm (#5602) · 8daacbd7
  milesial authored Feb 03, 2026
```
Signed-off-by: Alexandre Milesi <milesial@users.noreply.github.com>
```
  8daacbd7
29 Jan, 2026 1 commit

feat: Implement priority-based KV cache offload filtering (#5563) · 6271a31f

Yuewei Na authored Jan 29, 2026


Signed-off-by: Yuewei Na <248773860+nv-yna@users.noreply.github.com>
Signed-off-by: Yuewei Na <nv-yna@users.noreply.github.com>
Signed-off-by: yna <nv-yna@users.noreply.github.com>
Co-authored-by: Yuewei Na <nv-yna@users.noreply.github.com>

6271a31f

24 Jan, 2026 1 commit

chore(deps): bump vLLM to 0.14.0 (#5593) · 50f1e0e1

Alec authored Jan 23, 2026


Signed-off-by: alec-flowers <aflowers@nvidia.com>
Signed-off-by: Alec Flowers <aflowers@nvidia.com>
Signed-off-by: Alec <35311602+alec-flowers@users.noreply.github.com>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

50f1e0e1

23 Jan, 2026 2 commits
- fix: Fix race condition in TP>1 when ImmediateTransferResult arrives before CreateSlot (#5393) · 7fe89c74
  Kris Hung authored Jan 23, 2026
  
  7fe89c74
- fix: enabling cuda mem pools for vectorized transfer in kvbm (#5475) · cde3b2a5
  Olga Andreeva authored Jan 22, 2026
```
Signed-off-by: Olga Andreeva <124622579+oandreeva-nv@users.noreply.github.com>
```
  cde3b2a5
22 Jan, 2026 1 commit
- chore: update nixl to 0.9.0 (#5528) · 6dba119d
  Nate Mailhot authored Jan 22, 2026
  
  6dba119d
13 Jan, 2026 1 commit
- chore: Revert KVBM v2 transfer integration (#5406) · 9ca2923d
  jthomson04 authored Jan 13, 2026
  
  9ca2923d
12 Jan, 2026 2 commits
- fix: Synchronizing on new thread to avoid delaing TRTLLM (#5333) · 329752d9
  Kyle McGill authored Jan 12, 2026
  
  329752d9
- fix: GB200 mem allocation failure (#5328) · e3c3a6b9
  Qi Wang authored Jan 12, 2026
  
  e3c3a6b9
02 Jan, 2026 1 commit
- chore: update all copyright headers in repo to 2026 (#5130) · cf433e68
  Tushar Sharma authored Jan 02, 2026
```
Signed-off-by: Tushar Sharma <tusharma@nvidia.com>
```
  cf433e68
24 Dec, 2025 2 commits
- feat: KVBM Object Support - add runtime vars for object (#5063) · 096699c4
  Patrick authored Dec 23, 2025
```
Signed-off-by: Patrick Riel <priel@nvidia.com>
```
  096699c4
- feat: KVBM Object Support - add to storage types (#5060) · 853151d9
  Patrick authored Dec 23, 2025
```
Signed-off-by: Patrick Riel <priel@nvidia.com>
```
  853151d9
18 Dec, 2025 1 commit
- fix: Don't recursively drop blocks in KVBM (#4995) · da59c6ae
  jthomson04 authored Dec 18, 2025
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
  da59c6ae
10 Dec, 2025 1 commit
- feat: KVBM V2 optimized bounce buffer transfer + benchmark (#3947) · 01819b87
  jthomson04 authored Dec 09, 2025
```
Signed-off-by: jthomson04 <jwillthomson19@gmail.com>
```
  01819b87
03 Dec, 2025 1 commit
- fix: Updates CUDA fatbin for KVBM (#4700) · 991d8c72
  Olga Andreeva authored Dec 03, 2025
```
Signed-off-by: Olga Andreeva <oandreeva@nvidia.com>
```
  991d8c72
02 Dec, 2025 1 commit
- feat: Extend the KV Event Consolidator to support TensorRT-LLM (#4533) · 6067c59d
  Kris Hung authored Dec 02, 2025
```
Signed-off-by: krishung5 <krish@nvidia.com>
```
  6067c59d
01 Dec, 2025 1 commit
- fix: Page-align O_DIRECT writes for Lustre compatibility (#4508) · 5d9fc858
  Olga Andreeva authored Dec 01, 2025
```
Signed-off-by: Olga Andreeva <oandreeva@nvidia.com>
```
  5d9fc858
21 Nov, 2025 1 commit
- chore: Make nats_client private at crate level, various tidy up (#4513) · f05f7629
  Graham King authored Nov 21, 2025
```
Signed-off-by: Graham King <grahamk@nvidia.com>
```
  f05f7629