From 099c04feb1a69e7572d660a76397f318cb5799e4 Mon Sep 17 00:00:00 2001 From: FujitsuPolycom <87842395+FujitsuPolycom@users.noreply.github.com> Date: Sat, 29 Aug 2026 23:35:12 -0500 Subject: [PATCH] Explain copy-on-write prefix reuse Describe immutable row tails, authenticated page deltas, cross-root row-trunk coalescing, and SparkCache CUDA restore and placement ownership in the interactive explorer. Add README discoverability and a GPU-free HTML contract, and refresh deployment source digests because the test is part of the deployable SparkCache tree. Compatibility: documentation and regression coverage only. CacheIdentity wire values, digest salts, 256-token geometry, stored formats, and runtime behavior are unchanged. Cache namespace impact: none. Validation: 750 SparkCache tests passed with 7 skipped; 108 deployment tests passed with 1 skipped; Ruff passed; desktop 1440x900 and mobile 390x844 browser checks passed with no horizontal overflow or console errors. --- README.md | 3 + deploy/deepseek_v4/tp4_profile.json | 2 +- deploy/glm52_35bpw/profile.json | 2 +- docs/sparkcache-prefix-explainer.html | 107 +++++++++++++++++---- sparkcache/test_prefix_explainer.py | 133 ++++++++++++++++++++++++++ 5 files changed, 225 insertions(+), 22 deletions(-) create mode 100644 sparkcache/test_prefix_explainer.py diff --git a/README.md b/README.md index b947429..4193a0e 100644 --- a/README.md +++ b/README.md @@ -25,6 +25,9 @@ Python-placement restore. The source deployment separately qualifies native [the public image record](deploy/glm53_flash/IMAGE_ANNOUNCEMENT.md), [the GLM-5.3 validation](GLM53_FLASH_DFLASH7_LIVE_VALIDATION.md), and [the SparkCache CUDA restore record](GLM53_NATIVE_RESTORE_PERFORMANCE_VALIDATION.md). +The [interactive prefix-reuse explorer](docs/sparkcache-prefix-explainer.html) +shows logical boundaries, verified selection, copy-on-write publication, shared +row trunks, and bounded GPU-prefix attachment. The public image does not contain the model checkpoints. SparkCache CUDA restore and shared GPU-prefix qualification belong to a later source-bound runtime that has diff --git a/deploy/deepseek_v4/tp4_profile.json b/deploy/deepseek_v4/tp4_profile.json index e1f1dbe..0764ff2 100644 --- a/deploy/deepseek_v4/tp4_profile.json +++ b/deploy/deepseek_v4/tp4_profile.json @@ -4,7 +4,7 @@ "cache_model_profile": "deepseek-v4-fp8-hma", "published_runtime_base": "ghcr.io/fujitsupolycom/gb10-vllm-serving@sha256:6fc26fdad81a18f0fff67ce0a05f6d90165625ea2e1cac8a6f39bfb462017028", "sparkcache": { - "source_sha256": "48e008ba0cbd12f1ffae1c28388ea83310f41c6219c955e13d63ab171290d8de" + "source_sha256": "335296b16304ef9b13d4e3b5b68dce41d332dcd2abce8df2ae2a25451471a06a" }, "model": { "repository": "deepseek-ai/DeepSeek-V4-Flash-0731", diff --git a/deploy/glm52_35bpw/profile.json b/deploy/glm52_35bpw/profile.json index 91dcfe7..e740ca2 100644 --- a/deploy/glm52_35bpw/profile.json +++ b/deploy/glm52_35bpw/profile.json @@ -5,7 +5,7 @@ "published_runtime_base": "ghcr.io/fujitsupolycom/gb10-vllm-serving@sha256:6fc26fdad81a18f0fff67ce0a05f6d90165625ea2e1cac8a6f39bfb462017028", "base_image_requirement": "exact GLM-5.2 3.5-bpw R7 image recorded by the source container inspection", "sparkcache": { - "source_sha256": "48e008ba0cbd12f1ffae1c28388ea83310f41c6219c955e13d63ab171290d8de" + "source_sha256": "335296b16304ef9b13d4e3b5b68dce41d332dcd2abce8df2ae2a25451471a06a" }, "model": { "repository": "brandonmusic/GLM-5.2-EXL3-TR3v4-3.5bpw-MTP78", diff --git a/docs/sparkcache-prefix-explainer.html b/docs/sparkcache-prefix-explainer.html index 165a18a..2b18d9a 100644 --- a/docs/sparkcache-prefix-explainer.html +++ b/docs/sparkcache-prefix-explainer.html @@ -3,7 +3,7 @@
- +