Parallel image computation in clusters with task-distributor

SpringerPlus

Table 2 Runtime of the sequential and parallel parts of task-distributor in the cluster of Raspberry Pi computers when a single one, two, four or eight nodes (processors) are used

	Resolution	1st seq. part (s)	2nd seq. part (s)	Parallel part (s)
1 node used	200 × 150	0.210	0.204	4.472
	400 × 300	0.215	0.202	10.505
	800 × 600	0.206	0.194	34.918
	1600 × 1200	0.206	0.205	132.840
	3200 × 2400	0.242	0.233	609.427
	6400 × 4800	0.219	0.468	2434.630
2 nodes used	200 × 150	0.200	0.449	4.628
	400 × 300	0.199	0.680	7.603
	800 × 600	0.207	1.552	21.167
	1600 × 1200	0.217	4.665	73.481
	3200 × 2400	0.203	15.243	331.059
	6400 × 4800	0.211	152.037	1321.640
4 nodes used	200 × 150	0.198	0.545	4.984
	400 × 300	0.207	0.777	7.043
	800 × 600	0.203	1.632	14.945
	1600 × 1200	0.243	4.625	48.780
	3200 × 2400	0.230	15.146	209.375
	6400 × 4800	0.219	120.324	821.244
8 nodes used	200 × 150	0.216	0.741	6.115
	400 × 300	0.201	1.002	7.422
	800 × 600	0.242	1.839	12.727
	1600 × 1200	0.204	4.802	32.458
	3200 × 2400	0.218	15.543	131.780
	6400 × 4800	0.209	120.471	509.707

All values in the table are rounded to three decimal places behind the decimal point