Implementation of Parallel K-Means Algorithm to Estimate Adhesion Failure in Warm Mix Asphalt

Table 13

Computation of speedup (k = 10) for multiple Hadoop nodes v/s C-sequential & C + OpenMP.

Sample	Execution time using C-Sequential (sec)	Execution time using C + OpenMP (sec)	Execution time using 4 nodes (sec)	Speedup (C + OpenMP/4 nodes)	Speedup (sequential-C/4 nodes)
Dry dataset
WP170	161.08	127.68	103.67	1.23	1.55
HP180	161.15	126.49	98.85	1.27	1.63
HO180	161.12	125.4	99.27	1.26	1.62
WO150	160.96	126.8	102.96	1.23	1.56
WO160	161.88	125.2	98.81	1.26	1.63
WO170	155.92	126.81	98.85	1.28	1.57
WP160	162.37	125.2	96.45	1.29	1.68
WP150	161.24	128.1	101.01	1.26	1.59
1FT dataset
WP170	160.62	113.45	86.12	1.31	1.86
HP180	161.82	127.67	98.19	1.30	1.64
HO180	161.61	116.98	92.13	1.26	1.75
WO150	160.98	120.33	95.27	1.26	1.68
WO160	161.30	125.88	99.33	1.26	1.62
WO170	162.29	126.64	98.55	1.28	1.64
WP160	164.26	127.31	103.37	1.23	1.58
WP150	161.18	113.71	86.30	1.31	1.86
3FT dataset
WP170	145.04	111.27	83.57	1.33	1.73
HP180	160.71	125.23	94.911	1.31	1.69
HO180	162.28	125.06	95.21	1.31	1.70
WO150	162.21	117.03	89.23	1.31	1.81
WO160	162.36	125.98	96.19	1.30	1.68
WO170	161.13	126.4	94.93	1.33	1.69
WP160	161.89	120.06	90.59	1.32	1.78
WP150	162.30	117.17	95.14	1.23	1.70