Integrantes de equipo:                                                                      Universidad Tecnológica De La Mixteca

Bacilio Anota Marvin Salvador

Enriquez Pacheco Blanca Nieves

Robles Robles Emmanuel Guadalupe

Objetivo general: 

Contrastar el desempeño de la paralelización hecha con directivas OMP versus paralelización hecha con Threads, tomando como caso de estudio el procesamiento digital de imágenes al transformarlas de RGB a escala de grises. 

Implementación en Windows:

Paralelización del método fromRGBtoGRAY (promedios)
	código original (s/cambios)		2 hilos	4 hilos	6 hilos
WINDOWS	17.178908 ms	hilos	8.96935 ms SpeedUp 1.915290183	9.89895 ms SpeedUp 1.735427293	11.21620816 ms SpeedUp 1.531015643
		directivas OMP	7.69551 ms SpeedUp 2.232328722	8.105408 ms SpeedUp 2.119437788	7.575238 ms SpeedUp 2.267771389

Todos los tiempos están en milisegundos ms.

Original (ms)	2 threads	4 threads	6 threads	OMP 2 threads	OMP 4 threads	OMP 6 threads
18.4852	8.464	9.0665	9.9128	7.3246	9.4871	7.2372
25.2992	8.7255	9.0876	9.8575	7.0716	8.1517	7.3572
20.0945	8.4053	9.9743	14.2712	8.5979	8.9652	8.4074
17.8828	8.6084	9.1927	9.5909	7.3877	9.3376	7.4144
17.4528	8.4966	9.0894	13.1696	8.7314	7.2581	7.5824
18.3062	8.5747	11.2571	13.4694	6.7111	9.2042	8.0796
17.0709	9.0089	9.9721	13.0782	6.8964	7.4242	7.3414
17.5256	8.9114	10.9311	12.0645	8.7206	7.3314	7.9896
17.9173	9.2275	9.4992	13.1925	7.2144	7.6012	7.1208
17.2236	10.3646	9.6664	11.6222	7.0964	8.2748	6.8378
16.6896	8.3653	9.7365	10.5901	8.805	7.8347	7.1773
17.3892	8.9811	8.9942	11.1176	8.8646	7.9702	7.0997
17.4252	9.0713	9.9623	10.0163	11.1647	7.928	6.9256
16.8316	8.6696	9.5171	12.4181	7.1036	9.2296	7.3684
16.3214	9.4872	9.7274	10.4255	9.0493	8.9601	7.7848
16.6601	8.8657	9.0465	11.9686	10.4259	9.6391	8.3364
16.5353	8.7442	10.1977	12.003	6.8945	8.823	7.3011
17.0922	8.8114	9.699	9.7974	6.8347	8.5981	7.3089
16.578	8.7692	8.8335	9.1969	9.0062	7.3743	7.2026
16.1333	8.7707	9.2032	9.8478	8.6362	6.6153	7.7598
16.5682	8.7275	9.7208	13.6353	6.9028	7.5056	7.5795
16.6482	8.9619	11.3818	13.0545	6.7652	7.0465	9.7839
17.6035	9.1309	13.1845	9.2039	6.8987	7.7089	7.01
16.7093	8.4897	11.948	13.0885	7.4146	7.5918	7.2562
18.4033	8.635	8.9997	13.5645	8.8623	7.9713	6.9836
16.4101	8.8773	12.8058	9.484	7.1173	7.4258	7.5517
15.6351	9.1194	9.4191	13.0618	8.7051	7.8232	7.1332
15.8925	8.5103	10.2428	9.0417	7.0358	9.1803	7.5963
15.7044	8.5126	10.5257	13.2275	8.7172	8.988	7.459
16.7255	8.3779	11.2773	9.739	7.3673	7.8243	8.0069
16.0867	8.5944	9.6516	13.3236	6.9182	7.5338	9.1684
16.5878	9.5782	11.174	13.1312	6.9962	7.6767	7.2068
16.7808	9.3728	10.2892	12.8408	8.9522	8.9509	8.3071
17.8855	10.0437	9.1169	9.5453	6.7525	8.2593	7.6101
16.8956	8.7921	9.4433	9.2887	6.7581	7.453	7.7869
17.1536	9.0501	10.4227	8.8292	6.7471	7.2919	7.1268
16.6848	8.9506	9.2092	9.4132	6.8477	8.976	7.2119
17.1847	8.886	8.8156	9.4969	7.0691	8.4274	7.1486
16.7573	8.806	9.536	10.5482	8.9998	7.4805	7.3562
16.5339	8.3538	9.2858	10.5023	6.8185	7.2306	7.3254
16.3977	8.3758	10.0771	10.3657	8.9197	9.2819	7.5852
17.4225	8.9028	9.4533	14.164	7.0384	7.7542	7.1274
17.9068	8.7754	9.0712	9.2237	9.0104	9.1897	8.1585
16.7845	8.5328	10.3568	10.2611	7.0519	7.4486	7.444
16.7231	8.8499	8.8474	10.6651	6.8855	7.7907	8.0639
17.4818	8.911	8.8487	9.5466	7.0035	7.3869	8.264
16.3352	11.5138	9.8806	9.4725	6.9252	7.7104	6.8991
16.6482	8.7163	11.0203	13.5637	7.1035	8.753	8.3117
16.6417	12.3651	8.9865	10.7016	6.8998	7.8572	7.3452
16.8391	8.4318	9.302	11.4356	6.7551	7.7441	7.322

Implementación en Linux

Paralelización del método fromRGBtoGRAY (promedios)
	código original (s/cambios)		2 hilos	4 hilos	6 hilos
LINUX	92.02875137 milisegundos	Hilos	58.68086338 SpeedUp 1.568292388	39.86300945 SpeedUp 2.30862528	35.86229801 SpeedUp 2.566169947
		OMP	59.07876015 SpeedUp 1.557729904	39.10441876 SpeedUp 2.353410543	44.00159359 SpeedUp 2.091486782

Original	2 threads	4 threads	6 threads	OMP 2 threads	OMP 4 threads	OMP 6 threads
91.8200016	62.63399124	40.99297523	30.6198597	62.27993965	42.31595993	40.02809525
89.04600143	61.83981895	39.40296173	40.19284248	61.79904938	36.39698029	38.07806969
90.19184113	61.83600426	40.54808617	33.33497047	59.32283401	37.25886345	37.63198853
95.19505501	59.17906761	57.92403221	36.16404533	57.42502213	42.87099838	75.17004013
98.68788719	58.2818985	46.90694809	32.05704689	55.48477173	46.03719711	44.93093491
91.61114693	57.46102333	39.8709774	31.86392784	57.87205696	38.43688965	58.72011185
94.55084801	57.47509003	35.87603569	30.94696999	55.53197861	36.54193878	35.49313545
96.07195854	55.79400063	37.53685951	32.93395042	58.51101875	38.04397583	45.64809799
102.4968624	60.68921089	37.28508949	39.07608986	59.48400497	38.50102425	70.2650547
91.89987183	56.14113808	37.30797768	39.71505165	59.38100815	39.91389275	42.47903824
95.90387344	56.3519001	37.34111786	46.55885696	63.11297417	40.11702538	41.87011719
93.79816055	57.44409561	36.12709045	42.58012772	60.50491333	40.33184052	47.0969677
91.89796448	55.25708199	40.56811333	46.03600502	58.95900726	34.08908844	37.62388229
89.26510811	57.93309212	40.61102867	46.7441082	57.17802048	40.40598869	55.90105057
93.71709824	57.2218895	37.50896454	30.45797348	56.3352108	46.19407654	36.56506538
99.98512268	55.17292023	41.13292694	40.9719944	56.4558506	34.35516357	46.72598839
86.50708199	59.63397026	40.76910019	40.64321518	62.04319	38.21992874	34.02996063
91.36986732	60.92500687	39.34502602	31.25286102	59.92794037	39.9491787	35.31193733
90.99006653	76.77006721	39.43490982	29.98185158	58.09402466	34.91902351	37.59503365
88.52100372	65.73987007	39.1600132	41.32008553	63.22193146	39.65091705	43.68686676
88.79089355	67.83509254	39.64591026	31.92615509	57.48295784	39.59488869	48.81095886
93.64295006	59.54408646	40.43889046	35.38298607	58.97593498	37.82200813	45.00007629
89.32495117	56.00214005	40.09509087	31.88800812	60.69207191	40.41719437	47.22690582
89.75410461	58.23802948	42.57583618	32.48310089	59.30495262	39.28518295	51.97095871
93.66488457	56.90097809	40.54808617	31.62503242	63.0209446	37.00709343	43.03789139
89.89500999	55.94611168	38.79904747	34.94405746	58.38990211	38.17510605	50.02784729
89.87092972	61.79499626	41.06712341	33.70404243	57.22308159	38.39111328	36.23199463
90.42596817	54.63385582	39.33691978	36.34595871	59.70907211	40.78912735	38.80500793
89.76316452	58.69913101	43.25008392	38.98692131	64.2747879	36.01694107	40.73119164
92.08703041	59.23295021	41.50915146	39.14999962	61.00201607	44.21901703	47.38712311
91.01200104	58.98404121	39.7670269	30.98702431	62.51096725	41.15605354	45.25899887
88.14096451	59.95988846	38.79308701	33.6329937	59.43489075	38.89608383	53.76887321
91.63308144	55.93705177	42.07205772	34.46602821	59.09490585	39.91889954	40.30203819
90.65389633	57.02710152	38.42902184	31.65602684	58.70509148	39.1368866	41.04304314
95.61800957	58.58397484	35.88891029	38.78188133	56.66017532	37.00304031	45.12405396
88.36913109	56.661129	39.0920639	32.92489052	59.31305885	36.43918037	49.70097542
91.45689011	55.08780479	39.76106644	48.13599586	56.70690536	42.09303856	41.5148735
91.32313728	56.82611465	40.99392891	30.3030014	59.06796455	35.85481644	35.82596779
92.6759243	59.95106697	38.5658741	37.68014908	59.53383446	38.16008568	46.50497437
90.23809433	56.25915527	35.79497337	35.92085838	56.43010139	40.67707062	38.91086578
91.3131237	58.97688866	38.57707977	35.90393066	57.19995499	34.49583054	54.14485931
90.89398384	59.88907814	43.07699203	33.58817101	57.13415146	42.47307777	39.73197937
91.28594398	54.99196053	38.73205185	27.81701088	57.40499496	40.9450531	42.77300835
91.76397324	58.16984177	36.24200821	33.61797333	60.28199196	38.02609444	35.08090973
95.24011612	59.37385559	38.39898109	37.88495064	57.62982368	39.10422325	40.50183296
91.02392197	55.2418232	41.06497765	34.53302383	54.91781235	37.55784035	41.03207588
90.45696259	58.04300308	39.87812996	38.7070179	61.17796898	43.65706444	36.20696068
89.96987343	58.74896049	39.86501694	32.09400177	60.50300598	37.4019146	35.92705727
93.26791763	57.50703812	38.6068821	37.71495819	62.13784218	38.17510605	43.34592819
94.35391426	55.2148819	36.63396835	36.87691689	55.09209633	37.78195381	39.29901123

Diagrama de cajas y bigote.

Las cajas van ordenadas de izquierda a derecha por Windows original

 hasta Linux con directivas OMP

• Windows (Hilos):

• El tiempo de ejecución disminuye significativamente al usar 2 hilos en comparación con el código original.

• Añadir más hilos (4 y 6) aumenta ligeramente el tiempo de ejecución, lo que indica que el paralelismo adicional no proporciona beneficios y podría introducir sobrecarga.

• Windows (Directivas OMP):

• El uso de directivas OMP resulta en tiempos de ejecución consistentemente bajos, con poca variación entre 2, 4 y 6 hilos.

• Esto sugiere que OMP es eficiente en la gestión del paralelismo en este caso.

• Linux (Hilos):

• Se observa una disminución drástica en el tiempo de ejecución al pasar del código original a 2 hilos.

• Aumentar a 4 hilos reduce aún más el tiempo, pero usar 6 hilos resulta en un tiempo ligeramente mayor, similar al comportamiento observado en Windows.

• Linux (Directivas OMP):

• Al igual que en Windows, OMP ofrece un rendimiento consistente en Linux, con tiempos de ejecución bajos y poca variación entre diferentes números de hilos.

• Sin embargo, los tiempos de ejecución en Linux son notablemente más altos que en Windows, lo que podría deberse a que en Linux se está ejecutando en una máquina virtual (VMWARE).

Observaciones Generales:

• Tanto en Windows como en Linux, el uso de hilos ofrece una mejora significativa en el rendimiento en comparación con el código original.

• Las directivas OMP son una alternativa eficiente para gestionar el paralelismo, especialmente cuando se utilizan más de 2 hilos.

• La elección entre hilos y OMP dependerá de factores como la plataforma, el hardware y la facilidad de implementación.

•