[Lane 2 · native] NotImplemented stubs — BLAS acceleration on the JVM CPU backend #1425

Sub-issue of #1414 ([Feature]: Close the remaining NotImplemented stubs outside the operator surface).

Lane: 2 · Kotlin core
Skill needed: Kotlin JVM; FFM (Panama) or a BLAS binding; benchmarking.
Size: m (1–2 days)
Blocked by: nothing

What to do

1. PlatformCpuOpsFactory.jvm.kt notes "BLAS acceleration not yet implemented" and always uses DefaultCpuOps.
2. Prototype an FFM binding to a system BLAS sgemm behind the existing kernel-provider mechanism (see skainet-backends/skainet-backend-native-cpu), with a fallback when no BLAS is present.
3. Compare against the Panama Vector matmul with the engine benchmarks before proposing it as default.

Acceptance

• Benchmark numbers posted
• No behaviour change when BLAS is absent
• Result reported back on the parent issue

Notes

If something here turns out to be wrong or much bigger than the size label, comment on this issue or open a specific bug instead of blocking the parent.

Coverage matrix: docs/modules/ROOT/pages/reference/ops-status-matrix.adoc · How lanes work: Issue taxonomy