Bitsliced outer bit-18 reset detector
The complete outer T1/T2 bit-18 hard reset can be predicted exactly with 16-lane truth masks and ripple carry from one inherited M4 baseline state.
Diagnostic work, exact algebra and local capabilities are not treated as end-to-end mining advantage.
What was tested?
The complete outer T1/T2 bit-18 hard reset can be predicted exactly with 16-lane truth masks and ripple carry from one inherited M4 baseline state.
Why the test is meaningful
Bitslicing represents a Boolean value over all 16 source children as one 16-bit mask. XOR, AND and carry operations then evaluate every lane in parallel while preserving exact semantics. A hard reset occurs only when both outer operands freeze to the same bit.
mask(f) = Σ_{x=0}^{15} f(x)·2ˣcarry_{i+1}=(u_i∧v_i)∨(u_i∧carry_i)∨(v_i∧carry_i)outer reset at 18 ⇔ T1[18] and T2[18] are equal constantsHow it was tested
Construct bitsliced masks for inherited M4 components, evaluate exact ripple carries into bit 18, compile T1/T2 freeze and value outputs, and validate against direct per-child SHA arithmetic on original and fresh cohorts. Record a Python stage-count benchmark only as a cost signal.
What happened
Original and fresh semantic errors were zero. The detector found 545 outer hard resets in 1,361 original records and 111 in 300 fresh records. Component constancy was high, including 1,275 original and 285 fresh constant Σ0 bit-18 masks.
Exactness and statistical controls
All 16 lanes were checked against the direct oracle. Fresh law and detector errors were zero. The representation benchmark was explicitly barred from being interpreted as native CPU, ASIC or energy economics.
What the result means
The bitsliced detector is an exact executable specification of the outer reset. It improves representation and auditability, but its practical cost relative to a native hash implementation is not established.
Limitations
- Python stage counts are not hardware benchmarks.
- Truth-mask operations may map differently to CPU, GPU or ASIC logic.
- The reset is still only one internal event in a full candidate path.
Evidence trail
Recreate masks from the inherited baseline, evaluate every ripple stage, compare all 16 predicted lanes with direct children and report both semantic errors and actual detector cost separately.
Canonical variants
SUBENGINE-V28ASUBENGINE-V28BSUBENGINE-V28CSUBENGINE-V28DSUBENGINE-V28ESource: internally audited canonical reports. Local filesystem structure, private headers and operational identifiers are excluded from publication.