← Back to the chapter · Previous detail

Technical detail · Hardware Accelerators for Edge AI

Keep the decision record tied to evidence

13.11 Implementation Sketch

The sketch gives two points each for supported operators, sufficient memory headroom, measured latency within the deadline, and both thermal validation and rollback. Its returned risks make missing evidence visible; a high score does not waive a failed requirement.

def score_candidate(candidate, requirements):
    score = 0
    risks = []

    if candidate.supports(requirements.model_operators):
        score += 2
    else:
        risks.append("unsupported model operators")

    if candidate.memory_headroom >= requirements.memory_headroom:
        score += 2
    else:
        risks.append("insufficient memory headroom")

    if candidate.measured_latency_ms <= requirements.deadline_ms:
        score += 2
    else:
        risks.append("misses end-to-end latency deadline")

    if candidate.thermal_test_passed and candidate.update_path.has_rollback:
        score += 2
    else:
        risks.append("thermal or update validation incomplete")

    return {"candidate": candidate.name, "score": score, "risks": risks}

13.11.1 Evidence Ledger

CheckEvidence To RecordFailure Signal
Operator placementRuntime/compiler logs for the converted graphUnsupported layer or CPU fallback
Compute demandMACs per inference times inferences per secondMissed deadline despite good peak metrics
Memory headroomWeights, activations, input buffers, tensor arena, runtimeAllocation failure, swapping, or smaller batch/window than required
Sustained operationPower, thermals, and enclosure run under expected duty cycleThrottling, resets, battery miss, or heat limit
Lifecycle supportModel/runtime update path, monitoring, rollbackManual updates or unobservable fleet state

← Back to the chapter · Previous detail