EXP-062NEGATIVE
Canonical GPU engineering / SHA-256d

Compact Gray recurrence with constant deltas

A non-unrolled Gray loop with constant nonce and σ₀ deltas preserves reuse without the code expansion observed in EXP-137.

Claim statusNO END-TO-END MINING ADVANTAGE DEMONSTRATED

Diagnostic work, exact algebra and local capabilities are not treated as end-to-end mining advantage.

Date2026-08-14
HardwareNVIDIA GeForce RTX 5080 · exact SHA-256d verifier
Run scopePreregistered campaign with sealed result and audit
ReproducibilityBuild the frozen compact loop, verify the delta recurrence and repeat equal-work balanced timing.
01 / Question & hypothesis

What was tested?

A non-unrolled Gray loop with constant nonce and σ₀ deltas preserves reuse without the code expansion observed in EXP-137.

02 / Scientific basis

Why the test is meaningful

A compact recurrence isolates whether bookkeeping cost, rather than unrolling, consumes the exact linear reuse.

nonce_{i+1}=nonce_i⊕Δ_iσ₀_{i+1}=σ₀_i⊕σ₀(Δ_i)SASS_candidate≈SASS_baseline
Compact Gray recurrence with constant deltasVisual reading of the published metrics and gates for EXP-062; it summarizes the registered result, not a mining advantage.EXP-062 / COMPACT GRAY RECURRENCE WITH CONSTANT DELTASKERNEL RATIO0.99974495% INTERVAL0.998857–1.000632POSITIVE HEADERS1 / 4
FIGURE / RESULT READINGVisual reading of the published metrics and gates for EXP-062; it summarizes the registered result, not a mining advantage.
03 / Method

How it was tested

Use two 32-entry constant delta tables, explicitly prevent loop unrolling and rerun the four-header, 32-process balanced benchmark.

04 / Observed result

What happened

Kernel ratio0.999744
95% interval0.998857–1.000632
Positive headers1 / 4

SASS remained compact at 2,509 versus 2,507 instructions and exactness gates passed. Performance was neutral/slightly negative: kernel ratio 0.999744, 95% interval 0.998857–1.000632, positive on 1/4 headers.

05 / Validation

Exactness and statistical controls

Coverage, B32 hit, hashlib comparison, spills and register count all passed before timing was interpreted.

06 / Interpretation

What the result means

Both compact and unrolled Gray σ₀ reuse fail robust performance gates; bookkeeping consumes the theoretical reuse.

Limitations

  • Closure applies to the two registered implementations.
  • A different architecture may behave differently.
  • No probability-per-hash change is implied.
07 / Reproduction

Evidence trail

Build the frozen compact loop, verify the delta recurrence and repeat equal-work balanced timing.

Canonical variants

CANONICAL-EXP-138PREREGISTERED-CAMPAIGNSEALED-AUDIT

Source: internally audited canonical reports. Local filesystem structure, private headers and operational identifiers are excluded from publication.