Jmeter benchmark:Apisix VS Nginx, CPU occupancy rate huge?

[Git4Future]

We run another Nginx to listen 81 port,and it returns statics files response like 'helo.html'
this is my route config:
{ "uri": "/*", "name": "local-body", "methods": [ "GET", "POST" ], "host": "local-body.test.cn", "upstream": { "nodes": [ { "host": "10.0.1.94", "port": 81, "weight": 1 } ], "timeout": { "connect": 6, "send": 6, "read": 6 }, "type": "roundrobin", "scheme": "http", "pass_host": "pass", "keepalive_pool": { "idle_timeout": 75, "requests": 10000, "size": 1000 } }, "status": 1 }
Environment:
APISIX: 10.0.1.98 4C
NGINX: 10.0.1.19. 4C
JMETER:8C
NGINX UPSTREAM: 10.0.1.94 4C

This is the result from apisix:

This is the result from nginx:

why CPU occupancy rate between Apisix and Nginx are so huge?

[tzssangglass]

You can make a CPU flamegraph to see where the CPU consumption is