chore(release): prepare 0.21.0 - #566
Merged
Merged
Conversation
- gradle.properties: drop -SNAPSHOT; RELEASE_SIGNING_ENABLED stays true. - CHANGELOG: add 0.21.0 section covering the JVM Vector half of M5 (kernel SPI + Panama FP32 + tile-blocking + production routing + ServiceLoader auto-discovery + Q4_K SIMD + sibling SPI + Q4_K MemSeg + Q6_K SIMD + Q4_0 partial SIMD), plus ScratchPool SPI, TensorOps.permute, and Q4_K/Q5_K canonical layout fix. - README: bump Quickstart coordinates to 0.21.0; compact "What's New" section. - NATIVE_FFM_KERNEL_PROVIDER.md: PRD for the deferred priority-100 native FFM kernel provider — module layout, FFM binding pattern, staged delivery plan, success metrics, risks. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Removes NATIVE_FFM_KERNEL_PROVIDER.md from the release as a tracked file, plus its references in README's "What's New" and the CHANGELOG intro paragraph for the M5 section. The deferred native FFM kernel provider is still mentioned by name in the CHANGELOG (since the deferral is the actual release-relevant fact), just without a link to a doc that's no longer in the tree. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
3 tasks
MacOS
pushed a commit
to MacOS/SKaiNET
that referenced
this pull request
Jul 10, 2026
Three new explanation pages under docs/.../explanation/perf/ covering the M5 work that landed in 0.21.0: - simd-kernels.adoc — kernel SPI overview, FloatVector + FMA pattern, tile blocking, ServiceLoader auto-discovery + factory wrappers, KernelMatmulBench numbers (8.6×–10.8× over scalar at 256/512/1024 on Apple Silicon NEON). - quantized-simd-kernels.adoc — per-format pipelines (Q4_0, Q4_K, Q4_K MemSeg, Q6_K, Q8_0): ByteVector → AND/LSHR nibble extract → castShape(B2F) → fused FMA, plus the lazy-dmin trick and the per- format coverage matrix. - native-ffm-plan.adoc — recovers the FFM PRD content from git history (61962de:NATIVE_FFM_KERNEL_PROVIDER.md, dropped from 0.21.0 release per SKaiNET-developers#566). Module layout, FFM binding pattern, staged delivery, success metrics, risks, trigger conditions. architecture.adoc grows from a 4-line stub to an arc42-style reference with focus on the 0.21.0 changes: - Building Block View — module table + kernel SPI ASCII diagram (commonMain api + jvmMain auto-discovery + jvmMain providers). - Runtime View — eager-execution flow from `ctx.ops.matmul` through `chooseQuantizedMatmul` / `chooseMatmul` to the SPI kernel, with the lazy provider resolution and fall-through pattern called out. - Architecture decisions table, quality requirements, risks (Vector API still incubator, no native provider yet, prior reverts). nav.adoc: register the three new explanation pages under .Explanation. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Release prep for 0.21.0 — the JVM Vector half of milestone M5 (CPU backend dispatch). 10 PRs landed since 0.20.0:
After this release every quantized format in
JvmQuantizedVectorKernels(Q4_0, Q4_K, Q4_K MemSeg, Q6_K, Q8_0) is SIMD'd to some degree. FP32 matmul through the SPI hits 8.6×–10.8× over scalar at 256/512/1024 (Apple Silicon NEON, JMH); Q4_K matmul-vector hits ≈30/55/73 GFLOPS at 1024²/4096×1024/4096².Deferred
The priority-100 native (FFM) kernel provider that closes M5's literal
≥2.5× for Q4_Kmilestone metric is captured as a PRD inNATIVE_FFM_KERNEL_PROVIDER.md— module layout, FFM binding pattern, staged delivery plan, success metrics, risks. Trigger conditions for un-deferring documented at the bottom of the PRD.Files changed
gradle.properties—VERSION_NAME0.21.0-SNAPSHOT → 0.21.0CHANGELOG.md— new 0.21.0 sectionREADME.md— Quickstart coordinates 0.20.0 → 0.21.0; compact "What's New"NATIVE_FFM_KERNEL_PROVIDER.md(new) — PRD for the deferred native providerTest plan
./gradlew :skainet-backends:skainet-backend-cpu:compileKotlinJvm— clean../gradlew buildof a downstream consumer once 0.21.0 is published.🤖 Generated with Claude Code