MAYAN_ALFA — 100M SUMMARY BENCHMARK MANIFEST
============================================

Document role
-------------
This manifest documents the execution of the 100M summary benchmark layer inside
the MAYAN_ALFA P0.7 comparison workflow.

Project identity
----------------
MAYAN_ALFA is an Independent Computational Observation Framework focused on
measured benchmark behavior, deterministic comparison, scaling interpretation,
ARM64 computational behavior and publication-safe validation records.

Benchmark configuration
-----------------------
Scale: 100M
Limit: 100000000
Step: 1000000
Detail mode: 0
Output root: OUTPUT_100M
Execution timestamp: 2026-05-17 21:17:30 CEST
Working directory: [LOCAL_PROJECT_PATH_REDACTED] TESTOVÁNÍ/P0_7_MAYAN_ALFA_MR_PRIMESIEVE_RESET01

Engines / paths executed
------------------------
1. MAYAN_ALFA observation path
   Runner: P0_7A_MAYAN_ONLY/run_custom.sh
   Output: OUTPUT_100M/MAYAN

2. Miller–Rabin reference path
   Runner: P0_7B_MR_ONLY/run_custom.sh
   Output: OUTPUT_100M/MR

3. primesieve reference path
   Runner: P0_7C_PRIMESIEVE_ONLY/run_primesieve_p0_7.sh
   Output: OUTPUT_100M/PRIMESIEVE

Why this layer matters
----------------------
The 100M summary layer is a public-safe scaling checkpoint. It extends beyond the
10M detailed diagnostic layer while keeping output size controlled and suitable
for release packaging, QA review and publication documentation.

Interpretation boundary
-----------------------
The results are benchmark observations produced under a specific hardware,
software, compiler, dataset and workflow context. They support reproducible
comparison and scaling interpretation. They do not constitute a universal
mathematical proof, cryptographic guarantee or unrestricted performance claim.

Public/private boundary
-----------------------
This layer is intended to remain public-safe after standard leak-check and
sanitization. It must not include protected core source logic, private heuristics,
Observation Registry internals, commercial private workflows or controlled layer [CONTROLLED_SCALE_REDACTED]
validation archive data.

Completion
----------
Status: completed
Completion timestamp: 2026-05-17 21:17:55 CEST

Generated folders
-----------------
OUTPUT_100M
OUTPUT_100M/MAYAN
OUTPUT_100M/MR
OUTPUT_100M/PRIMESIEVE

Generated files
---------------
OUTPUT_100M/MAYAN/MAYAN_LIMIT_100000000_STEP_1000000.txt
OUTPUT_100M/MAYAN/MAYAN_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260510_060109.csv
OUTPUT_100M/MAYAN/MAYAN_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260510_112941.csv
OUTPUT_100M/MAYAN/MAYAN_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260515_022355.csv
OUTPUT_100M/MAYAN/MAYAN_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260515_023038.csv
OUTPUT_100M/MAYAN/MAYAN_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260517_211730.csv
OUTPUT_100M/MAYAN/output_100M_ALL_RUN_WALL.txt
OUTPUT_100M/MAYAN/output_MAYAN_100M_WALL.txt
OUTPUT_100M/MR/MR_LIMIT_100000000_STEP_1000000.txt
OUTPUT_100M/MR/MR_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260510_060109.csv
OUTPUT_100M/MR/MR_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260510_112941.csv
OUTPUT_100M/MR/MR_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260515_022355.csv
OUTPUT_100M/MR/MR_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260517_211731.csv
OUTPUT_100M/PRIMESIEVE/output_PRIMESIEVE_100M.txt
OUTPUT_100M/PRIMESIEVE/PRIMESIEVE_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260510_113005.csv
OUTPUT_100M/PRIMESIEVE/PRIMESIEVE_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260515_022419.csv
OUTPUT_100M/PRIMESIEVE/PRIMESIEVE_SUMMARY_LIMIT_0100000000_STEP_0001000000_20260517_211754.csv
OUTPUT_100M/RUN_100M_SUMMARY_ALL_MANIFEST.txt
OUTPUT_100M/RUN_100M_SUMMARY_ALL_PUBLIC_NOTE.txt

Recommended next validation steps
---------------------------------
1. Run the compare/validation layer after all required public benchmark scales
   are available.
2. Include this 100M manifest in audit and publication workpacks.
3. Confirm that public release packages do not contain protected source logic,
   private registry materials, controlled layer [CONTROLLED_SCALE_REDACTED] numeric archives or internal-only
   workflow data.
4. Use bounded language in public documentation: observation, measurement,
   comparison, scaling behavior and validated workflow context.

