Best Use & Engineering Application Guide

Engineering Category: Deterministic Sovereign Compute Kernel

Best Use in Production: Pure-Rust C64-bounded integer transformation kernel verified across 64 canonical test vectors for deterministic Snap-Circuit composition.

When to Select This Capsule: Verified baseline member of a 2-capsule Equivalence Class; useful as a reference or cross-validation peer. Low-latency L1-resident kernel (4018219/400000 cyc/op); well-suited for high-frequency pipeline stages.

Snap-Circuit Wiring Role: Specialized Multi-Pin Stage `(u64,u64,u64)->u64` — wire matching input pins via `snap_dag` or `snap_pipeline`.

Hardware _rdtsc: 4018219/400000 cycles/op
Latency: 649379/200000 ns/op
Throughput: 200000000/649379 ops/us
Compiled ELF64 .text: 125 B (33 x86_64 insns)
C64 64KB RAM Budget: 525/65536 (525 B source)
Ternary Vote: + (score 1) | Salience: 426835/5304
Pure-Rust Capsule Source Code (525 B)
pub fn aurora_mopd2_distill_loss_q15(student_q15: u64, mimo_teacher_q15: u64, audio8_teacher_q15: u64) -> u64 {
    let diff_mimo = if student_q15 >= mimo_teacher_q15 { student_q15 - mimo_teacher_q15 } else { mimo_teacher_q15 - student_q15 };
    let diff_audio8 = if student_q15 >= audio8_teacher_q15 { student_q15 - audio8_teacher_q15 } else { audio8_teacher_q15 - student_q15 };
    let weighted = (diff_mimo.saturating_mul(60) + diff_audio8.saturating_mul(40)) / 100;
    if weighted > 32767 { 32767 } else { weighted }
}
rustc -C opt-level=3 x86_64 Disassembly (125 B .text / 33 insns)
aurora_mopd2_distill_loss_q15:
    mov    rcx, rdx
    mov    rax, rdi
    sub    rax, rsi
    sub    rsi, rdi
    cmovae    rax, rsi
    mov    rdx, rdi
    sub    rdx, rcx
    sub    rcx, rdi
    cmovb    rcx, rdx
    mov    edx, 60
    mul    rdx
    jo    .LBB7_2
    mov    rsi, rax
    mov    edx, 40
    mov    rax, rcx
    mul    rdx
    jo    .LBB7_4
.LBB7_5:
    add    rax, rsi
    shr    rax, 2
    movabs    rcx, 2951479051793528259
    mul    rcx
    shr    rdx, 2
    cmp    rdx, 32767
    mov    eax, 32767
    cmovb    rax, rdx
    ret
.LBB7_2:
    mov    rsi, -1
    mov    edx, 40
    mov    rax, rcx
    mul    rdx
    jno    .LBB7_5
.LBB7_4:
    mov    rax, -1
    jmp    .LBB7_5

Head-to-Head Functional Equivalence Class Peers (2 Implementation(s) Matching 64 Test Vectors)

Global RankCapsule & DomainHardware _rdtscLatencyThroughputELF .textTernary VoteAuthor Aye
#51st_golf_aurora_mopd2_distill_loss_q15_u64
domain: aurora-mopd2 | id: 78104ea09e7dab2d19ada6c506ae3858
291299/40000 cyc/op790293/400000 ns/op400000000/790293 ops/us125 B .text (33 insns)+ (1)aye::Hue+Gemini(0+1)::x86_64 @ 432Hz
#520aurora_mopd2_distill_loss_q15 (THIS CAPSULE)
domain: signal | id: d437ee3acb7491f7c3e1a081301eaf73
4018219/400000 cyc/op649379/200000 ns/op200000000/649379 ops/us125 B .text (33 insns)+ (1)aye-aurora-8b-voice @ 432Hz