KENNING AI OPTIMIZATION

EDGE AI PERFORMANCE BENCHMARKING v1.0

KENNING ONLINE
AI MODELS READY

AI MODEL STATUS

YOLO v8

Object Detection

OPTIMIZED
Inference: 12.3ms

ResNet-50

Image Classification

OPTIMIZED
Inference: 8.7ms

SegNet

Lane Detection

OPTIMIZED
Inference: 15.2ms

MobileNet

Vehicle Detection

OPTIMIZED
Inference: 6.1ms

INFERENCE TIME COMPARISON

MEMORY USAGE

PLATFORM COMPARISON

NVIDIA JETSON
Average Inference
10.5ms
Power Consumption
15.2W
CPU Usage
45%
NXP i.MX 8M
Average Inference
28.3ms
Power Consumption
8.5W
CPU Usage
72%
RISC-V
Average Inference
42.7ms
Power Consumption
5.8W
CPU Usage
85%

CPU/GPU UTILIZATION

POWER CONSUMPTION

OPTIMIZATION LOG