Skip to content

标签:Edge AI

IRQ Affinity Tuning: Spread Network and Storage Interrupts

Linux lands most interrupts on CPU0 by default, leaving one core saturated while others idle and total throughput capped. Interrupts defaulting to CPU0 is the trap Inspect: cat /proc/interrupts shows which core each IRQ lands on — watch NIC and storage controllers; Pin them: set /proc/irq//smp_affinity to spread NIC interrupts across the A76 cores; Pair […]

阅读全文 →

Machine-Vision Inspection Lines: From Camera Choice to Cycle Time

Many teams pick a model and then discover it won’t keep up. The right order: throughput → per-frame budget → required compute → part number. Compute takt before compute power Reverse-engineer takt: at 60 parts/minute across 4 stations each inference gets roughly 200ms — reserve half for capture and post-processing; Cameras and buses: GigE Vision […]

阅读全文 →

Converting YOLO to RKNN: From ONNX to Running on the NPU

Most RKNN conversion failures trace to toolchain/driver version mismatch. Align versions first, then argue about the model. Version matching is the first gate Version triangle: PC rknn-toolkit2, on-device rknn_server/runtime and the kernel RKNPU driver must be mutually compatible; Operator check: read the verbose build output to confirm nothing falls back to CPU; Output alignment: RKNN […]

阅读全文 →

CPU Frequency Governor Tuning: Balancing Speed, Power and Heat

Default governors may not match your workload — sustained full load and bursty work want completely different policies. Understand the governors first Policy choice: performance suits sustained load but runs hot; ondemand or schedutil suits bursts; powersave fits always-idle devices; Big.LITTLE scheduling: RK3588 has 4x A76 plus 4x A55 — pin realtime work to A76 […]

阅读全文 →