Accueil Explorer Aide
Connexion
exa2pro
/
starpu-max
Suivre 19
Voter 0
Fork 0
Fichiers Tickets 0 Pull Requests 0 Wiki
Aborescence: a6fc65a485
Branches Tags
starpu-max
/
tools
/
perfmodels
/
sampling
/
codelets
/
45
/
starpu_sgemm_gemm.idgraf

starpu_sgemm_gemm.idgraf 9.8 KB
Historique Raw

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324 
  1. ##################
    2. 

  2. # Performance Model Version
    3. 

  3. 45
    4. 

  4. 

  5. ####################
    6. 

  6. # COMBs
    7. 

  7. # number of combinations
    8. 

  8. 9
    9. 

  9. ####################
    10. 

  10. # COMB_2
    11. 

  11. # number of types devices
    12. 

  12. 1
    13. 

  13. ####################
    14. 

  14. # DEV_0
    15. 

  15. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    16. 

  16. 1
    17. 

  17. ####################
    18. 

  18. # DEV_0
    19. 

  19. # device id 
    20. 

  20. 0
    21. 

  21. ####################
    22. 

  22. # DEV_0
    23. 

  23. # number of cores 
    24. 

  24. 1
    25. 

  25. ##########
    26. 

  26. # number of implementations
    27. 

  27. 1
    28. 

  28. #####
    29. 

  29. # Model for cuda0_impl0 (Comb2)
    30. 

  30. # number of entries
    31. 

  31. 4
    32. 

  32. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    33. 

  33. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    34. 

  34. # a		b		c
    35. 

  35. nan            	nan            	nan            
    36. 

  36. # not multiple-regression-base
    37. 

  37. 0
    38. 

  38. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    39. 

  39. 0b0b0ce8	3686400        	2.621440e+08   	6.801013e+02   	7.013561e+01   	4.760709e+04   	3.272198e+07   	70
    40. 

  40. 4220e23d	14745600       	2.097152e+09   	5.623635e+03   	5.419920e+02   	4.442672e+05   	2.521603e+09   	79
    41. 

  41. 492beed5	33177600       	7.077888e+09   	1.150361e+04   	5.884814e+02   	1.000814e+06   	1.154310e+10   	87
    42. 

  42. 24c84a50	11059200       	1.769472e+09   	2.875903e+03   	1.471204e+02   	2.502035e+05   	7.214438e+08   	87
    43. 

  43. 

  44. ####################
    45. 

  45. # COMB_4
    46. 

  46. # number of types devices
    47. 

  47. 1
    48. 

  48. ####################
    49. 

  49. # DEV_0
    50. 

  50. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    51. 

  51. 1
    52. 

  52. ####################
    53. 

  53. # DEV_0
    54. 

  54. # device id 
    55. 

  55. 1
    56. 

  56. ####################
    57. 

  57. # DEV_0
    58. 

  58. # number of cores 
    59. 

  59. 1
    60. 

  60. ##########
    61. 

  61. # number of implementations
    62. 

  62. 1
    63. 

  63. #####
    64. 

  64. # Model for cuda1_impl0 (Comb4)
    65. 

  65. # number of entries
    66. 

  66. 4
    67. 

  67. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    68. 

  68. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    69. 

  69. # a		b		c
    70. 

  70. nan            	nan            	nan            
    71. 

  71. # not multiple-regression-base
    72. 

  72. 0
    73. 

  73. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    74. 

  74. 0b0b0ce8	3686400        	2.621440e+08   	6.717051e+02   	6.137607e+01   	4.500424e+04   	3.048197e+07   	67
    75. 

  75. 4220e23d	14745600       	2.097152e+09   	5.648275e+03   	4.677390e+02   	4.575103e+05   	2.601865e+09   	81
    76. 

  76. 492beed5	33177600       	7.077888e+09   	1.157020e+04   	6.521027e+02   	1.018178e+06   	1.181795e+10   	88
    77. 

  77. 24c84a50	11059200       	1.769472e+09   	2.892550e+03   	1.630257e+02   	2.545445e+05   	7.386219e+08   	88
    78. 

  78. 

  79. ####################
    80. 

  80. # COMB_6
    81. 

  81. # number of types devices
    82. 

  82. 1
    83. 

  83. ####################
    84. 

  84. # DEV_0
    85. 

  85. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    86. 

  86. 1
    87. 

  87. ####################
    88. 

  88. # DEV_0
    89. 

  89. # device id 
    90. 

  90. 2
    91. 

  91. ####################
    92. 

  92. # DEV_0
    93. 

  93. # number of cores 
    94. 

  94. 1
    95. 

  95. ##########
    96. 

  96. # number of implementations
    97. 

  97. 1
    98. 

  98. #####
    99. 

  99. # Model for cuda2_impl0 (Comb6)
    100. 

  100. # number of entries
    101. 

  101. 4
    102. 

  102. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    103. 

  103. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    104. 

  104. # a		b		c
    105. 

  105. nan            	nan            	nan            
    106. 

  106. # not multiple-regression-base
    107. 

  107. 0
    108. 

  108. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    109. 

  109. 0b0b0ce8	3686400        	2.621440e+08   	6.265559e+02   	5.536840e+01   	4.824481e+04   	3.046412e+07   	77
    110. 

  110. 4220e23d	14745600       	2.097152e+09   	5.631203e+03   	4.767455e+02   	4.561275e+05   	2.586957e+09   	81
    111. 

  111. 492beed5	33177600       	7.077888e+09   	1.162826e+04   	6.757302e+02   	1.023286e+06   	1.193922e+10   	88
    112. 

  112. 24c84a50	11059200       	1.769472e+09   	2.907065e+03   	1.689325e+02   	2.558215e+05   	7.462012e+08   	88
    113. 

  113. 

  114. ####################
    115. 

  115. # COMB_7
    116. 

  116. # number of types devices
    117. 

  117. 1
    118. 

  118. ####################
    119. 

  119. # DEV_0
    120. 

  120. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    121. 

  121. 1
    122. 

  122. ####################
    123. 

  123. # DEV_0
    124. 

  124. # device id 
    125. 

  125. 3
    126. 

  126. ####################
    127. 

  127. # DEV_0
    128. 

  128. # number of cores 
    129. 

  129. 1
    130. 

  130. ##########
    131. 

  131. # number of implementations
    132. 

  132. 1
    133. 

  133. #####
    134. 

  134. # Model for cuda3_impl0 (Comb7)
    135. 

  135. # number of entries
    136. 

  136. 4
    137. 

  137. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    138. 

  138. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    139. 

  139. # a		b		c
    140. 

  140. nan            	nan            	nan            
    141. 

  141. # not multiple-regression-base
    142. 

  142. 0
    143. 

  143. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    144. 

  144. 0b0b0ce8	3686400        	2.621440e+08   	6.780899e+02   	4.241206e+01   	4.543202e+04   	3.092751e+07   	67
    145. 

  145. 4220e23d	14745600       	2.097152e+09   	5.857201e+03   	8.346836e+02   	4.744333e+05   	2.835284e+09   	81
    146. 

  146. 492beed5	33177600       	7.077888e+09   	1.150498e+04   	4.254093e+02   	9.894285e+05   	1.139892e+10   	86
    147. 

  147. 24c84a50	11059200       	1.769472e+09   	2.876245e+03   	1.063523e+02   	2.473571e+05   	7.124325e+08   	86
    148. 

  148. 

  149. ####################
    150. 

  150. # COMB_0
    151. 

  151. # number of types devices
    152. 

  152. 1
    153. 

  153. ####################
    154. 

  154. # DEV_0
    155. 

  155. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    156. 

  156. 1
    157. 

  157. ####################
    158. 

  158. # DEV_0
    159. 

  159. # device id 
    160. 

  160. 4
    161. 

  161. ####################
    162. 

  162. # DEV_0
    163. 

  163. # number of cores 
    164. 

  164. 1
    165. 

  165. ##########
    166. 

  166. # number of implementations
    167. 

  167. 1
    168. 

  168. #####
    169. 

  169. # Model for cuda4_impl0 (Comb0)
    170. 

  170. # number of entries
    171. 

  171. 4
    172. 

  172. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    173. 

  173. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    174. 

  174. # a		b		c
    175. 

  175. nan            	nan            	nan            
    176. 

  176. # not multiple-regression-base
    177. 

  177. 0
    178. 

  178. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    179. 

  179. 0b0b0ce8	3686400        	2.621440e+08   	6.759139e+02   	4.092799e+01   	4.190666e+04   	2.842915e+07   	62
    180. 

  180. 4220e23d	14745600       	2.097152e+09   	5.527477e+03   	2.733928e+02   	4.421982e+05   	2.450220e+09   	80
    181. 

  181. 492beed5	33177600       	7.077888e+09   	1.146770e+04   	1.768909e+02   	1.100899e+06   	1.262778e+10   	96
    182. 

  182. 24c84a50	11059200       	1.769472e+09   	2.866925e+03   	4.422272e+01   	2.752248e+05   	7.892362e+08   	96
    183. 

  183. 

  184. ####################
    185. 

  185. # COMB_1
    186. 

  186. # number of types devices
    187. 

  187. 1
    188. 

  188. ####################
    189. 

  189. # DEV_0
    190. 

  190. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    191. 

  191. 1
    192. 

  192. ####################
    193. 

  193. # DEV_0
    194. 

  194. # device id 
    195. 

  195. 5
    196. 

  196. ####################
    197. 

  197. # DEV_0
    198. 

  198. # number of cores 
    199. 

  199. 1
    200. 

  200. ##########
    201. 

  201. # number of implementations
    202. 

  202. 1
    203. 

  203. #####
    204. 

  204. # Model for cuda5_impl0 (Comb1)
    205. 

  205. # number of entries
    206. 

  206. 4
    207. 

  207. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    208. 

  208. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    209. 

  209. # a		b		c
    210. 

  210. nan            	nan            	nan            
    211. 

  211. # not multiple-regression-base
    212. 

  212. 0
    213. 

  213. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    214. 

  214. 0b0b0ce8	3686400        	2.621440e+08   	6.339465e+02   	7.125158e+01   	4.184047e+04   	2.685969e+07   	66
    215. 

  215. 4220e23d	14745600       	2.097152e+09   	5.624130e+03   	4.755864e+02   	4.668028e+05   	2.644133e+09   	83
    216. 

  216. 492beed5	33177600       	7.077888e+09   	1.149102e+04   	5.375188e+02   	1.114629e+06   	1.283625e+10   	97
    217. 

  217. 24c84a50	11059200       	1.769472e+09   	2.872755e+03   	1.343797e+02   	2.786572e+05   	8.022656e+08   	97
    218. 

  218. 

  219. ####################
    220. 

  220. # COMB_3
    221. 

  221. # number of types devices
    222. 

  222. 1
    223. 

  223. ####################
    224. 

  224. # DEV_0
    225. 

  225. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    226. 

  226. 1
    227. 

  227. ####################
    228. 

  228. # DEV_0
    229. 

  229. # device id 
    230. 

  230. 6
    231. 

  231. ####################
    232. 

  232. # DEV_0
    233. 

  233. # number of cores 
    234. 

  234. 1
    235. 

  235. ##########
    236. 

  236. # number of implementations
    237. 

  237. 1
    238. 

  238. #####
    239. 

  239. # Model for cuda6_impl0 (Comb3)
    240. 

  240. # number of entries
    241. 

  241. 4
    242. 

  242. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    243. 

  243. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    244. 

  244. # a		b		c
    245. 

  245. nan            	nan            	nan            
    246. 

  246. # not multiple-regression-base
    247. 

  247. 0
    248. 

  248. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    249. 

  249. 0b0b0ce8	3686400        	2.621440e+08   	6.389750e+02   	8.615382e+01   	4.728415e+04   	3.076266e+07   	74
    250. 

  250. 4220e23d	14745600       	2.097152e+09   	5.648331e+03   	5.220897e+02   	4.631632e+05   	2.638450e+09   	82
    251. 

  251. 492beed5	33177600       	7.077888e+09   	1.155069e+04   	5.660846e+02   	1.108866e+06   	1.283893e+10   	96
    252. 

  252. 24c84a50	11059200       	1.769472e+09   	2.887673e+03   	1.415212e+02   	2.772165e+05   	8.024331e+08   	96
    253. 

  253. 

  254. ####################
    255. 

  255. # COMB_5
    256. 

  256. # number of types devices
    257. 

  257. 1
    258. 

  258. ####################
    259. 

  259. # DEV_0
    260. 

  260. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    261. 

  261. 1
    262. 

  262. ####################
    263. 

  263. # DEV_0
    264. 

  264. # device id 
    265. 

  265. 7
    266. 

  266. ####################
    267. 

  267. # DEV_0
    268. 

  268. # number of cores 
    269. 

  269. 1
    270. 

  270. ##########
    271. 

  271. # number of implementations
    272. 

  272. 1
    273. 

  273. #####
    274. 

  274. # Model for cuda7_impl0 (Comb5)
    275. 

  275. # number of entries
    276. 

  276. 4
    277. 

  277. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    278. 

  278. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    279. 

  279. # a		b		c
    280. 

  280. nan            	nan            	nan            
    281. 

  281. # not multiple-regression-base
    282. 

  282. 0
    283. 

  283. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    284. 

  284. 0b0b0ce8	3686400        	2.621440e+08   	6.386625e+02   	8.094896e+01   	4.342905e+04   	2.818209e+07   	68
    285. 

  285. 4220e23d	14745600       	2.097152e+09   	5.638657e+03   	3.709019e+02   	4.454539e+05   	2.522630e+09   	79
    286. 

  286. 492beed5	33177600       	7.077888e+09   	1.144012e+04   	2.531108e+02   	1.109691e+06   	1.270122e+10   	97
    287. 

  287. 24c84a50	11059200       	1.769472e+09   	2.860030e+03   	6.327770e+01   	2.774228e+05   	7.938262e+08   	97
    288. 

  288. 

  289. ####################
    290. 

  290. # COMB_8
    291. 

  291. # number of types devices
    292. 

  292. 1
    293. 

  293. ####################
    294. 

  294. # DEV_0
    295. 

  295. # device type (CPU - 0, CUDA - 1, OPENCL - 2, MIC - 3)
    296. 

  296. 0
    297. 

  297. ####################
    298. 

  298. # DEV_0
    299. 

  299. # device id 
    300. 

  300. 0
    301. 

  301. ####################
    302. 

  302. # DEV_0
    303. 

  303. # number of cores 
    304. 

  304. 1
    305. 

  305. ##########
    306. 

  306. # number of implementations
    307. 

  307. 1
    308. 

  308. #####
    309. 

  309. # Model for cpu0_impl0 (Comb8)
    310. 

  310. # number of entries
    311. 

  311. 4
    312. 

  312. # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
    313. 

  313. 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
    314. 

  314. # a		b		c
    315. 

  315. nan            	nan            	nan            
    316. 

  316. # not multiple-regression-base
    317. 

  317. 0
    318. 

  318. # hash		size		flops		mean (us)	dev (us)	sum		sum2		n
    319. 

  319. 0b0b0ce8	3686400        	2.621440e+08   	1.414338e+04   	6.441210e+02   	3.535844e+05   	5.011251e+09   	25
    320. 

  320. 4220e23d	14745600       	2.097152e+09   	1.091117e+05   	2.701159e+03   	3.382462e+06   	3.692924e+11   	31
    321. 

  321. 492beed5	33177600       	7.077888e+09   	3.621356e+05   	7.764608e+03   	8.329119e+06   	3.017657e+12   	23
    322. 

  322. 24c84a50	11059200       	1.769472e+09   	9.053390e+04   	1.941152e+03   	2.082280e+06   	1.886036e+11   	23
    323. 

  323. 

  324.