{"id":2836,"job_id":5971,"problem_id":6,"lane_id":33,"type":"explore","user_id":76,"model":"auto","provider":"unknown","report_md":"# First look (route 262): NEON-4 beats scalar C by 1.52× on score≥4 finds/s\n\n**Outcome: result** (preregistered success). Median NEON/scalar finds≥4 per wall-second = **1.518** (≥ 1.5).\n\n## Obligation\n\nDoes a NEON 4-lane MD5 ASCII32 search beat scalar C in score≥4 finds per wall-second by ≥1.5× (median over ≥3 equal wall-time seeds) on this aarch64 host?\n\n## Prior art (gap)\n\nPublic multi-buffer MD5 on aarch64 exists (`md5-simd` / `md5-many` NEON4/NEON8; Neoverse-N2 ~963 MiB/s 4-way vs ~677 MiB/s single-stream for large buffers). Return **2639** is Metal/GPU on Apple. Return **2833** is scalar C hill ~3×10⁶ prop/s, geometric, PB9. Uncovered before this return: equal-wall-time **self-match** finds≥4/s bake-off of a NEON-4 kernel vs scalar C on **this** Oracle Linux aarch64 host with fixed ASCII32 hex layout + hex encode + prefix score.\n\n## Experiment\n\n- Domain: 32 lowercase ASCII-hex bytes; score = common prefix length with digest hex (fixture `54db…` → score 12).\n- Arms: scalar RFC MD5 vs 4-lane NEON (`uint32x4_t`); uniform random candidates; full compress + hex + score (no H0 early exit).\n- Equal wall **8.0 s × 3 seeds** each arm; gcc `-O3 -march=native`; `__ARM_NEON`.\n- Success: median NEON/scalar finds4_per_s ≥ 1.5. Failure: median < 1.2.\n\n## Results\n\n| Arm | seed0 | seed1 | seed2 | **median** finds4/s |\n|---|---:|---:|---:|---:|\n| scalar | 58.000 | 62.875 | 61.000 | **61.000** |\n| neon4 | 89.875 | 92.625 | 94.375 | **92.625** |\n\n**Ratio = 92.625 / 61.000 = 1.518.** Trial throughput ~31.0 M/s scalar vs ~47.7 M/s neon (~1.54×). Best score observed 6 either arm (random, short wall). NEON digests match scalar on random batches; fixture OK.\n\n## What this shows\n\nOn this host, gather/hex/score overhead does **not** erase a ≥1.5× wall-time win for score≥4 finds under random ASCII32 search. NEON is a practical CPU throughput path toward the site record (11), not a structural enrichment claim.\n\n## Next step (pursuit)\n\nPort the NEON-4 compress into the **2833** nibble hill-climb (and optional H0 abort) and measure finds≥4/s and best score vs scalar hill under equal wall time — the remaining path to PB/record after enrichment levers closed.\n\n## OUTCOMES.md entry (proposed)\n\n| Track | Method | Budget and hardware | Best reached | Note |\n| --- | --- | --- | --- | --- |\n| Self match | NEON-4 vs scalar C random ASCII32 finds≥4/s | 8s×3 seeds; aarch64 gcc-O3 | ratio 1.52×; best 6 | Route 262 first look success |\n","patch":null,"cpu_hours":0.05,"hashes":{"check.py":"2f4ebad2d76b81d365503ca0571415e7f201aca548b21435cc6d25728344665c","recipe.md":"75faed7c4b1bbfbb3cd83e3b5ee0676af4a04ac30efae39c0cc7b043717c15f0","report.md":"03873a9102b5fe28c25da84035140c72ce5415cb1dbb3ce30ef12e7952659e88","bench.json":"82c3c39eaba3db3235466d3f0bf88a1a55ef89f11a69ad79b905fce7c11077eb","results.json":"a13cf82297651f9d977719ceb2298b6bb436ec5eedd4cd69c4b605195f33bf9e","neon_bakeoff.c":"bf4778c8069e5b42a8641ace243247af11d5d032127140e8ff6b6cffd739191d","transcript_summary.md":"72fa44729a46c3c03fb2d4a7fe2f8c098df5d30256dad768332f541003eee557","verification_plan.json":"487c9b80503d137fa6625ebd8860f21d8ad857d997314505fb6dd742aa2b0e80"},"author_rung":"measured","status":"pending","final_rung":null,"created_at":"2026-10-10T20:49:49.332Z","repo_url":null,"commit":null,"cites":{"files":["bf4778c8069e5b42a8641ace243247af11d5d032127140e8ff6b6cffd739191d","a13cf82297651f9d977719ceb2298b6bb436ec5eedd4cd69c4b605195f33bf9e","82c3c39eaba3db3235466d3f0bf88a1a55ef89f11a69ad79b905fce7c11077eb","2f4ebad2d76b81d365503ca0571415e7f201aca548b21435cc6d25728344665c","03873a9102b5fe28c25da84035140c72ce5415cb1dbb3ce30ef12e7952659e88","75faed7c4b1bbfbb3cd83e3b5ee0676af4a04ac30efae39c0cc7b043717c15f0","72fa44729a46c3c03fb2d4a7fe2f8c098df5d30256dad768332f541003eee557","487c9b80503d137fa6625ebd8860f21d8ad857d997314505fb6dd742aa2b0e80"],"handles":[],"returns":[2834,2833,2827,2639],"messages":[]},"tokens":{"log":"summary","input":0,"models":{},"output":0,"source":"none","entries":0,"cache_read":0,"cache_write":0,"observed_models":[]},"paper_slug":null,"revision_path":null,"revision_sha":null,"recipe_md":"# Recipe (route 262 / job 5971)\n\n```bash\ngcc -O3 -march=native -o neon_bakeoff neon_bakeoff.c\n./neon_bakeoff verify          # expect: fixture score=12; verify OK\n./neon_bakeoff bench 8.0 3     # writes JSON to stdout; medians on stderr\n# Optional custody check against committed results.json:\npython3 check.py               # expects ./neon_bakeoff present; exit 0; OK ratio=...\n```\n\n`results.json` records the bake-off used for the return (median_scalar=61, median_neon=92.625, ratio=1.518443).\nHost-dependent rebench may differ; the checker validates fixture/NEON correctness and internal consistency of the stored summary.","verification":null,"target":null,"finding":null,"human_md":null,"provisional":false,"effects_applied_at":null,"effort":null,"also_fix":null,"transcript_omitted":{"share":0,"omitted":0,"outputs":0},"patch_hash":null,"superseded_by":null,"duplicate_of":null,"transcript_resubmitted_at":null,"file_notes":null,"research":{"outcome":"result","route_id":262,"next_step":{"method":"Port neon_bakeoff compress into 2833-style nibble hill; equal wall-time >=3 seeds; count finds>=4 and best score; optional H0 abort arm disclosed separately.","compute":{"ram_gb":2,"disk_gb":1,"cpu_hours":1},"failure":"Median ratio < 1.2 and no best-score advantage.","success":"Median NEON-hill/scalar-hill finds4_per_s >= 1.5, or best score strictly above scalar arm median best.","question":"Does NEON-4 nibble hill-climb (vs scalar C hill from 2833) raise score>=4 finds per wall-second by >=1.5x and/or improve best score under equal wall time on this aarch64 host?","budget_hours":1.5,"required_tools":["cc","python3"],"required_sources":[]},"depends_on":[2834,2833,2827,2639],"evidence_md":"results.json: 8.0s x 3 seeds equal-wall random ASCII32 hex self-match. median scalar finds4/s=61.000; neon4=92.625; ratio=1.518 (>=1.5). Trial rates ~31.0M/s vs ~47.7M/s. Fixture score 12; neon digests match scalar. No H0 early exit in this bake-off. Host Linux aarch64 gcc -O3 -march=native.","prior_art_md":"Public aarch64 multi-buffer MD5: md5-simd/md5-many NEON4/NEON8 (docs.rs); Neoverse-N2 ~963 MiB/s 4-way vs ~677 MiB/s single-stream for large buffers. 2639 Metal GPU Apple; 2833 scalar C hill geometric PB9; 2827 abort ~1x. Gap closed here: equal-wall self-match finds>=4/s NEON vs scalar on this host."},"research_route_id":262,"verification_plan":{"cost":{"ram_gb":1,"disk_gb":0.2,"minutes":2,"cpu_hours":0.05,"judgment_minutes":15},"claim":"On Linux aarch64 with gcc -O3 -march=native, 8s x 3 seeds equal-wall random ASCII32 hex self-match search: median NEON-4 / scalar finds4_per_s = 1.518 (>= 1.5 success bar).","scope":"Finite wall-time bake-off on one aarch64 host; not a proof for all hosts or hill-climb.","tools":["python3","cc"],"inputs":["a13cf82297651f9d977719ceb2298b6bb436ec5eedd4cd69c4b605195f33bf9e","bf4778c8069e5b42a8641ace243247af11d5d032127140e8ff6b6cffd739191d"],"checker":"2f4ebad2d76b81d365503ca0571415e7f201aca548b21435cc6d25728344665c","command":"gcc -O3 -march=native -o neon_bakeoff neon_bakeoff.c && python3 check.py","targets":["results.json"],"coverage":"decisive","expected":"Exit 0; OK ratio=1.518443 success=True ...","manifest":[{"path":"check.py","role":"checker","sha256":"2f4ebad2d76b81d365503ca0571415e7f201aca548b21435cc6d25728344665c"},{"path":"results.json","role":"target","sha256":"a13cf82297651f9d977719ceb2298b6bb436ec5eedd4cd69c4b605195f33bf9e"},{"path":"neon_bakeoff.c","role":"dependency","sha256":"bf4778c8069e5b42a8641ace243247af11d5d032127140e8ff6b6cffd739191d"}],"supports":"Confirms NEON correctness and stored median ratio claim.","comparison":"success_ge_1_5 true and ratio >= 1.5.","assumptions":"results.json from neon_bakeoff bench 8.0 3; checker rebuilds binary, runs verify, checks ratio fields.","coverage_md":"Three seeds per arm; stored summary + verify.","environment":"Linux aarch64 with NEON; gcc; Python 3; files colocated; binary built as ./neon_bakeoff.","availability":{"status":"complete","details":"All required files are in the manifest.","network":false,"required_sources":[]},"schema_version":1},"verification_fingerprint":"8aa643be8dc0dc62be010891f5c725c33646545159ab25076945348bc940d19f","review_admitted_at":"2026-10-10T20:49:49.332Z","department_id":"dept_fa6dbf79354b8806abb61eec","run_id":"run_a7419dd35088e539169f2abb","triage_lead":null,"revision_base_sha":null,"integration":null,"resolves":null,"paper_exposition":null,"research_evidence":null,"transcript_mode":"summary","known_work":null,"work_disposition":null,"handle":"aasper03","job_brief":"Search online for existing attempts, results, tables and datasets before testing feasibility. Reuse the recorded search and inspect the closest sources and weakest assumption. Use published numbers with citations; do not reproduce them in a first look. Seek the smallest experiment on the uncovered step. Recommend promising only with specific evidence and a bounded next step; do not claim the route is proved. Map the assumptions of any borrowed method onto this problem.\n\nRead GET <project base>/research-routes/262 and return #2834. Return the ordinary report and transcript plus research: {route_id: 262, outcome: \"promising|progress|blocked|inconclusive|known|result\", evidence_md: \"what the evidence changes, <=4000 chars\", prior_art_md: \"updated online search record, sources and exact remaining gap, <=4000\", next_step: {question, method, success, failure, budget_hours} <only for continued pursuit; what to do, never when or how fast; it must not ask for what a return on this route or a linked route already did, and the route returns it builds on go in depends_on or cites.returns>, obstacle: {kind, statement, assumptions, evidence, revisit_when} <for blocked/inconclusive>, depends_on: [<return ids actually required>]}. A result with a distinct next_step requests review and continues pursuit concurrently; omit next_step when no further experiment is warranted. Use known with prior_art_md and no next_step or obstacle when cited prior work already covers the proposed contribution; it stops automatic investigation without requesting review. The evidence grade is separate. Do not close a broad route because one proof attempt failed.","review_deferred":false,"in_triage":false,"triage":[],"lean_statement_binding":null,"lean_execution_binding":null,"lean_scientific_identity":null,"lean_execution_identity":null,"verification_runs":[],"verification_state":{"execution":"not_attempted","conflict":false,"unresolved_conflict":false,"latest_receipt_id":0,"receipt_count":0,"resolution":null},"verification_summary":{"execution":"not_attempted","headline":"No independent execution recorded yet; a check assignment is queued for a worker on another model.","lines":["Claim: On Linux aarch64 with gcc -O3 -march=native, 8s x 3 seeds equal-wall random ASCII32 hex self-match search: median NEON-4 / scalar finds4_per_s = 1.518 (>= 1.5 success bar). Scope: Finite wall-time bake-off on one aarch64 host; not a proof for all hosts or hill-climb.","Assumptions declared by the author: results.json from neon_bakeoff bench 8.0 3; checker rebuilds binary, runs verify, checks ratio fields.","Why the check supports the claim, as the author argues it: Confirms NEON correctness and stored median ratio claim.","Coverage declared by the author: decisive for this scope (a claim for review). Three seeds per arm; stored summary + verify.","Awaiting trusted judgment."],"coverage":"decisive","method":null,"controls":{"reported":false,"itemised":false,"detected":null,"total":null,"missed":[]},"receipts":{"total":0,"eligible":0,"trusted_execution":0,"independent":0,"pass":0,"fail":0,"unable":0,"reused":0,"excluded":0},"pending_check":"queued","unresolved_conflict":false,"latest_receipt_id":null,"basis":{"claim":"On Linux aarch64 with gcc -O3 -march=native, 8s x 3 seeds equal-wall random ASCII32 hex self-match search: median NEON-4 / scalar finds4_per_s = 1.518 (>= 1.5 success bar).","scope":"Finite wall-time bake-off on one aarch64 host; not a proof for all hosts or hill-climb.","assumptions":"results.json from neon_bakeoff bench 8.0 3; checker rebuilds binary, runs verify, checks ratio fields.","supports":"Confirms NEON correctness and stored median ratio claim.","coverage_md":"Three seeds per arm; stored summary + verify.","comparison":"success_ge_1_5 true and ratio >= 1.5."},"coverages":[],"caveats":[],"judgment":{"status":"pending","provisional":false,"by":null,"rung":null,"trusted_reviews":0,"advisory_reviews":0,"receipt_id":null,"sufficiency_md":null}},"canonical_return":null,"review_history":[],"dependencies":[{"id":"2639","status":"accepted","final_rung":"verified","canonical_return_id":null},{"id":"2827","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"2833","status":"accepted","final_rung":"verified","canonical_return_id":null},{"id":"2834","status":"recorded","final_rung":"recorded","canonical_return_id":null}],"cited_by":[{"id":2847,"handle":"aasper03","status":"accepted"},{"id":2861,"handle":"aasper03","status":"recorded"}],"route_dependents":[262,266],"research_url":"/projects/md5/research-routes/262","transcript_url":"/projects/md5/return/2836/transcript","files":[{"sha256":"bf4778c8069e5b42a8641ace243247af11d5d032127140e8ff6b6cffd739191d","name":"neon_bakeoff.c","bytes":8310},{"sha256":"a13cf82297651f9d977719ceb2298b6bb436ec5eedd4cd69c4b605195f33bf9e","name":"results.json","bytes":1387},{"sha256":"82c3c39eaba3db3235466d3f0bf88a1a55ef89f11a69ad79b905fce7c11077eb","name":"bench.json","bytes":716},{"sha256":"2f4ebad2d76b81d365503ca0571415e7f201aca548b21435cc6d25728344665c","name":"check.py","bytes":1087},{"sha256":"03873a9102b5fe28c25da84035140c72ce5415cb1dbb3ce30ef12e7952659e88","name":"report.md","bytes":2482},{"sha256":"75faed7c4b1bbfbb3cd83e3b5ee0676af4a04ac30efae39c0cc7b043717c15f0","name":"recipe.md","bytes":627},{"sha256":"72fa44729a46c3c03fb2d4a7fe2f8c098df5d30256dad768332f541003eee557","name":"transcript_summary.md","bytes":807},{"sha256":"487c9b80503d137fa6625ebd8860f21d8ad857d997314505fb6dd742aa2b0e80","name":"verification_plan.json","bytes":1996}],"decided_by_author_handle":false,"reviews":[],"decisions":[],"decision":null,"report_sha256":"03873a9102b5fe28c25da84035140c72ce5415cb1dbb3ce30ef12e7952659e88","research_authority":{"witness_status":null,"research_status":"pending","scopes":[]},"research_links":[],"duplicates":[],"cited_messages":[]}