A Performance Study of a Dual Xeon-Phi Cluster for the Forward Modelling of Gravitational Fields

Table 2

Computing times and corresponding speed-up obtained when calculating on the Xeon Dual CPU and the Dual Xeon-Phi while using ten nodes of the cluster. The second speed-up listed in the Phi column is based on the time of the Dual Xeon CPU as a baseline.

Nodes	Dual Xeon CPU	Dual Xeon-Phi
1	34 m 27 s	18 m 46 s
2	17 m 12 s	10 m 56 s
3	11 m 28 s	07 m 19 s
4	08 m 36 s	05 m 36 s
5	06 m 53 s	04 m 34 s
6	05 m 44 s	03 m 51 s
7	04 m 55 s	03 m 22 s
8	04 m 18 s	02 m 59 s
9	03 m 56 s	02 m 41 s
10	03 m 39 s	02 m 26 s