Parallel Implementation of FEM Solver for Shared Memory Using OpenMP

Table 12

Conjugate Gradient solver parallel timing and speedup for system of equations having 131585 unknowns.

Threads	1	2	3	4	5	6	7	8
CG	23.42	17.26	12.64	10.10	8.43	7.29	6.90	6.10
Speedup	1.00	1.36	1.85	2.32	2.78	3.21	3.39	3.85
Threads	9	10	11	12	13	14	15	16
CG	5.46	5.12	4.69	4.39	4.23	3.85	3.65	3.62
Speedup	4.29	4.57	5.00	5.34	5.54	6.09	6.42	6.47