This is an old revision of the document!

Lumerical benchmarks

solver speed result from logfile:

~> grep 'FDTD solver speed on process 0' x001/benchmark_small_p0_v241.log
    FDTD solver speed on process 0: 18.1763 Mnodes/s

Host	CPU	Cores/Node	v212	v222	v232	v241	Node@v241
x001	2x E5-2695v2	24	18.0825	18.091	18.1409	18.1763	436.232
x166	2x E5-2680v3	24	23.783	23.849	23.743	23.8678	572.827
xscale01	4x E7-8890v3	72	15.8967	15.738	14.9804	15.0902	1086.49
xgpu48	1x EPYC 7402P	24	34.6805	34.2354	31.9402	32.8264	787.833
xq02	2x EPYC 7713	128	14.8917	14.8524	14.6908	14.7221	1884.43