MAYAN_ALFA — NEON / SIMD ANALYSIS MANIFEST
==========================================

Script:
  01G_RUN_NEON_ANALYSIS.sh

Run timestamp UTC:
  2026-05-23 19:44:56 UTC

Working directory:
  [LOCAL_PROJECT_PATH_REDACTED] TESTOVÁNÍ/P0_7_MAYAN_ALFA_MR_PRIMESIEVE_RESET01

Purpose:
  This manifest documents the optional NEON / SIMD system-analysis layer of the
  MAYAN_ALFA P0.7 benchmark workflow. The layer records platform-oriented
  diagnostic observations related to ARM64 / NEON behavior and SIMD experiment
  probes. It supports benchmark interpretation, publication context and future
  optimization research.

Framework identity:
  MAYAN_ALFA is an Independent Computational Observation Framework. This means
  the project observes, measures, compares, archives and interprets bounded
  benchmark behavior under documented workflow conditions.

Interpretation boundary:
  The outputs of this layer are platform diagnostics and optimization context.
  They are not universal mathematical proof, not a guarantee of identical
  performance on other systems and not a replacement for independent
  reproduction.

Public/private boundary:
  Sanitized NEON/SIMD reports may support public GitHub, Zenodo and paper
  workpack materials when they do not expose protected source logic, internal
  registry content, private workflow data or protected [CONTROLLED_SCALE_REDACTED] datasets.

Expected output folders:
  OUTPUT_SYSTEM_ANALYSIS/NEON_BITMAP
  OUTPUT_SYSTEM_ANALYSIS/SIMD_NEON

Initial status:
  NEON bitmap scan: NOT_RUN
  SIMD NEON probe:  NOT_RUN

Final status:
  NEON bitmap scan: OK
  SIMD NEON probe:  OK

Generated / centralized reports:
  - OUTPUT_SYSTEM_ANALYSIS/NEON_BITMAP/P0_7E_NEON_BITMAP_SCAN_REPORT.txt
  - OUTPUT_SYSTEM_ANALYSIS/SIMD_NEON/SIMD_NEON_PROBE_REPORT.txt

Operational note:
  Missing NEON/SIMD experiment folders are not treated as fatal in publication
  workflow mode. This allows sanitized or partial release trees to remain
  buildable while still documenting which optional diagnostic layers were
  present.

Recommended next steps:
  1. Review generated reports for private paths or sensitive metadata.
  2. Run the public sanitation and leak-check workflow before publication.
  3. Keep [CONTROLLED_SCALE_REDACTED] validated datasets restricted to [CONTROLLED_ARCHIVE_REDACTED].
  4. Use this layer only as supporting system-analysis evidence.

