参考链接:Benchmark results
一. 基于RK3399/ARM-CPU的测试
1. 在本地PC上开启RPC Tracker
python3 -m tvm.exec.rpc_tracker
运行完显示如下:
INFO:root:If you are running ROCM/Metal, fork will cause compiler internal error. Try to launch with arg ```--no-fork```
INFO:RPCTracker:bind to 0.0.0.0:9190
2. 在RK3399上Register devices to the tracker
python3 -m tvm.exec.rpc_server --tracker=172.20.15.184:9190 --key=rk3399
运行完显示如下:
INFO:root:If you are running ROCM/Metal, fork will cause compiler internal error. Try to launch with arg ```--no-fork```
INFO:RPCServer:bind to 0.0.0.0:9090
3. 在本地PC上Verify the device registration We can query all registered devices by
python3 -m tvm.exec.query_rpc_tracker
运行完显示如下:
Tracker address localhost:9190
Server List
----------------------------
server-address key
----------------------------
172.20.15.65:34300 server:rk3399
----------------------------
Queue Status
------------------------------
key total free pending
------------------------------
rk3399 1 1 0
------------------------------
4. 在本地PC上Run benchmark
# ARM CPU
python3 arm_cpu_imagenet_bench.py --model rk3399 --rpc-key rk3399
# Mali GPU
# NOTE: To make the test environment more stable, we close GUI and lock the frequency
sudo /etc/init.d/lightdm stop
sudo -i
echo performance > /sys/class/misc/mali0/device/devfreq/ff9a0000.gpu/governor
python3 mobile_gpu_imagenet_bench.py --model rk3399 --rpc-key rk3399
python3 mobile_gpu_imagenet_bench.py --model rk3399 --rpc-key rk3399 --dtype float16