The INT8 Scale and the Pruned MAC Count
Ada re-derives this chapter’s own numbers step by step, at full precision
ADA · CALCULATION AUDIT
The INT8 Scale and the Pruned MAC Count
The chapter makes two optimization claims: an INT8 scale of 0.50 / 127 quantizes a 0.236 weight to the integer 60, and structured pruning cuts a 24-million-MAC inference to 18 million. Both must close before anyone calls the model “smaller.” This audit re-runs the INT8 scale and the pruned MAC count at full precision, then bounds exactly what the numbers justify.
Companion to the chapter Model Optimization for Edge AI — every number here comes from that chapter.
Ada: This chapter makes two numeric claims I want to close before anyone calls the optimized model “smaller”: that an INT8 scale of 0.50 / 127 quantizes a 0.236 weight to 60, and that structured pruning cuts a 24-million-MAC inference to 18 million. Let me re-run both at full precision.
- Scale and round-trip. The chapter rounds the scale to
0.00394for display; the stored value is0.50 / 127 = 0.00393701. Quantizing givesround(0.236 / 0.00393701) = round(59.944) = 60— the same integer either way. Reconstructing with the exact scale is60 x 0.00393701 = 0.23622, so this weight’s round-trip error is|0.236 - 0.23622| = 0.00022, comfortably below the worst-case half-step of0.00393701 / 2 = 0.00196850. - Weight-tensor storage.
2,400,000 weights x 4 bytes = 9,600,000 bytes = 9.6 MBin FP32;x 1 byte = 2.4 MBin INT8. The saving is9.6 - 2.4 = 7.2 MB, a4x(75%) cut on the weight tensor alone. - Pruned compute. Removing 25% of channels takes
24,000,000 x 0.75 = 18,000,000 MACs. At 5 inferences per second that is18,000,000 x 5 = 90,000,000 MACs/sinstead of24,000,000 x 5 = 120,000,000 MACs/s— a30,000,000 MACs/s, or 25%, reduction.
The audit closes the trade only for the weight tensor and the dense MAC count. A 4x file shrink and a 25% compute cut are real, but activation memory, calibration behaviour near the decision boundary, and the exported runtime graph are separate measurements — the numbers here justify continuing the conversion, not shipping it.
Every number above is taken from the chapter’s own material and re-derived step by step.