ViPar: High-Level Design Space Exploration for Parallel Video Processing Architectures

Table 5

Estimations for utilization, power, and frame execution time for pipe8 and pipe12 designs.

#		Freq.in MHz	Level of para-llelism	Slice (54650)	FF (437200)	LUT (218600)	BRAM (1090)	Power in mW	Frame exec. time in ms
25	Pipe8	100	1	15608	47992	48170	203	391.73	44.85
26			2	28745	87339	89700	387	510.4	24.838
27			3	41882	126686	131230	571	618.42	17.854
28			4	43190	166033	172760	755	649.25	14.32
29		150	1	16625	44339	48591	203	729.75	12.167
30			2	30357	80033	90533	387	650.15	16.596
31			3	44089	115727	132475	571	813.63	11.927
32			4	43605	151421	174417	755	837.69	9.565
33		200	1	16278	48644	53711	203	548.49	22.574
34			2	30001	88643	100770	387	777.02	12.494
35			3	43724	128642	147829	571	993.97	8.976
36			4	48722	168641	194888	755	1092.14	7.196
37	Pipe12	100	1	26073	67071	66804	275	475.8	30.965
38			2	49292	125497	126967	531	652.76	17.854
39			3	46783	183923	187130	787	674.48	12.774
40		150	1	24133	61240	69833	275	567.12	20.693
41			2	45392	113835	133017	531	818.27	11.927
42			3	49051	166430	196201	787	893.56	8.531
43		200	1	22127	67339	68978	275	646	15.581
44			2	41461	126033	131286	531	957.05	8.976
45			3	48399	184727	193594	787	1093.59	6.418