LabsPublished research, findings, and evidence.

Research

Measurements and published findings.

Research records document what FARPY tested, what the evidence showed, and where conclusions stop.

Studies

  • Octane Pixel Determinism Versus Runtime Instability

    Decoded pixels matched across runs; runtime variance failed the acceptance gate, so no performance result was published.

    Research · 2026-07-20 · Study · FRS-2026-001

    View study
  • Blender GPU Runtime Reproducibility Under a Frozen Workload Contract

    The frozen Blender contract passed artifact and scoring checks. V5 has not been physically executed.

    Research · 2026-07-20 · Study · FRS-2026-002

    View study
  • Reproducibility of a Seeded Bell-State Measurement Simulator

    A seeded Bell-state simulator reproduced byte-for-byte. No physical QPU was used.

    Research · 2026-07-29 · Study · FQR-2026-001

    View study

Research Notes

  • Why Identical Pixels Can Produce Different PNG File Hashes

    Yes, in the documented Octane experiment. The decoded pixel payloads matched across the measured outputs, while the complete PNG file hashes differed. The canonical evidence attributed the container-level byte differences to changing PNG metadata.

    Research · 2026-07-21 · Research Note · FRN-2026-001

    View study
  • Why Runtime Variance Can Invalidate an Otherwise Correct Benchmark

    Yes, under the tested Octane contract and configuration. The decoded pixel payloads matched, but timing did not satisfy the frozen acceptance rule. The runtime coefficient-of-variation gate was at most 3%. The best consecutive three-pass CV was 15.579055%, and additional warmups did not stabilize the measurements. FARPY therefore rejected performance publication and published no verified Octane benchmark result.

    Research · 2026-07-21 · Research Note · FRN-2026-002

    View study

Compatibility / Technical material

  • Blender 4.1.1 compatibility testing on FARPY RTX 3090 workers

    4 of 4 registered RTX 3090 GPUs qualified for isolated Blender 4.1.1 research. Production eligibility NOT_GRANTED.

    Research · 2026-07-29 · Compatibility study

    View study
  • FARPY Public Technical Brief

    Historical review-candidate public-model summary. Not an FRS result.

    Archived · 2026-07-22 · Technical Brief

    View study