Parallel Implementation of FEM Solver for Shared Memory Using OpenMP

Table 2

Speedup and efficiency of mesh read module using block distribution (525313 nodes and 1050628 elements).

Threads	1	2	3	4	5	6	7	8
Speedup	1.000	1.830	2.738	3.525	4.187	4.889	5.520	6.089
Efficiency	100.000	91.485	91.271	88.137	83.737	81.483	78.857	76.109
Threads	9	10	11	12	13	14	15	16
Speedup	6.651	7.146	7.584	8.140	8.310	8.803	9.100	9.107
Efficiency	73.895	71.465	68.943	67.837	63.921	62.881	60.667	56.917