Hello, GKHanna:
Would you please provide more details about your test case for ethernet latency?
As you mentioned ‘ping localhost’, I tried in my side.
To get similar result with PC, it’s better to boost TX1 CPU/EMMC frequency.
Generally, when system loading is low, those clocks will keep in low level to save power.
localhost ping result and tegrastats result will look like:
64 bytes from localhost (127.0.0.1): icmp_seq=1 ttl=64 time=0.075 ms
64 bytes from localhost (127.0.0.1): icmp_seq=2 ttl=64 time=0.590 ms
64 bytes from localhost (127.0.0.1): icmp_seq=3 ttl=64 time=0.765 ms
64 bytes from localhost (127.0.0.1): icmp_seq=4 ttl=64 time=0.802 ms
RAM 414/3853MB (lfb 782x4MB) cpu [0%,0%,0%,1%]@102 EMC 2%@68 AVP 2%@80 GR3D 0%@76 EDP limit 1734
RAM 414/3853MB (lfb 782x4MB) cpu [3%,0%,0%,2%]@102 EMC 0%@1065 AVP 2%@80 GR3D 0%@76 EDP limit 1734
RAM 414/3853MB (lfb 782x4MB) cpu [0%,0%,0%,1%]@102 EMC 2%@68 AVP 2%@80 GR3D 0%@76 EDP limit 1734
RAM 414/3853MB (lfb 782x4MB) cpu [0%,3%,0%,2%]@102 EMC 2%@68 AVP 2%@80 GR3D 0%@76 EDP limit 1734
RAM 414/3853MB (lfb 782x4MB) cpu [5%,0%,4%,2%]@102 EMC 2%@68 AVP 2%@80 GR3D 0%@76 EDP limit 1734
After boosting the system, the result will look like:
64 bytes from localhost (127.0.0.1): icmp_seq=43 ttl=64 time=0.045 ms
64 bytes from localhost (127.0.0.1): icmp_seq=44 ttl=64 time=0.041 ms
64 bytes from localhost (127.0.0.1): icmp_seq=45 ttl=64 time=0.038 ms
64 bytes from localhost (127.0.0.1): icmp_seq=46 ttl=64 time=0.038 ms
RAM 421/3853MB (lfb 769x4MB) cpu [0%,0%,0%,0%]@1734 EMC 0%@1600 AVP 2%@80 GR3D 0%@998 EDP limit 1734
RAM 421/3853MB (lfb 769x4MB) cpu [0%,0%,0%,0%]@1734 EMC 0%@1600 AVP 2%@80 GR3D 0%@998 EDP limit 1734
RAM 421/3853MB (lfb 769x4MB) cpu [1%,0%,0%,0%]@1734 EMC 0%@1600 AVP 3%@80 GR3D 0%@998 EDP limit 1734
RAM 421/3853MB (lfb 769x4MB) cpu [1%,0%,0%,0%]@1734 EMC 0%@1600 AVP 3%@80 GR3D 0%@998 EDP limit 1734
br
ChenJian