Browse Source

sgemm: Add perfmodels for pure square 1920 3D tiles

Samuel Thibault 4 years ago
parent
commit
3f9e0e111a

+ 8 - 4
tools/perfmodels/sampling/codelets/45/starpu_sgemm_gemm.attila

@@ -29,7 +29,7 @@
 #####
 # Model for cpu0_impl0 (Comb3)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -41,6 +41,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	3.328725e+05   	1.185902e+04   	2.563119e+07   	8.542747e+12   	77
 c00cf6b7	29491200       	7.077888e+09   	3.328725e+05   	1.185902e+04   	2.563119e+07   	8.542747e+12   	77
 78a2cc08	29491200       	7.077888e+09   	3.328725e+05   	1.185902e+04   	2.563119e+07   	8.542747e+12   	77
+a7cdf15b	44236800       	1.415578e+10   	6.657450e+05   	2.371804e+04   	5.126238e+07   	3.417099e+13   	77
 24c84a50	11059200       	1.769472e+09   	8.321812e+04   	2.964755e+03   	6.407798e+06   	5.339217e+11   	77
 0b0b0ce8	3686400        	2.621440e+08   	1.421718e+04   	3.409134e+02   	9.098993e+05   	1.294364e+10   	64
 4220e23d	14745600       	2.097152e+09   	1.008105e+05   	2.361630e+03   	8.064841e+06   	8.134670e+11   	80
@@ -67,7 +68,7 @@ c00cf6b7	29491200       	7.077888e+09   	3.328725e+05   	1.185902e+04   	2.56311
 #####
 # Model for cuda0_impl0 (Comb0)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -79,6 +80,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.123499e+04   	6.785566e+01   	1.190909e+06   	1.338033e+10   	106
 c00cf6b7	29491200       	7.077888e+09   	1.123499e+04   	6.785566e+01   	1.190909e+06   	1.338033e+10   	106
 78a2cc08	29491200       	7.077888e+09   	1.123499e+04   	6.785566e+01   	1.190909e+06   	1.338033e+10   	106
+a7cdf15b	44236800       	1.415578e+10   	2.246998e+04   	1.357113e+02   	2.381818e+06   	5.352132e+10   	106
 24c84a50	11059200       	1.769472e+09   	2.808747e+03   	1.696392e+01   	2.977272e+05   	8.362706e+08   	106
 0b0b0ce8	3686400        	2.621440e+08   	6.738679e+02   	4.393713e+01   	6.873452e+04   	4.651489e+07   	102
 4220e23d	14745600       	2.097152e+09   	5.557425e+03   	3.241733e+02   	5.835297e+05   	3.253957e+09   	105
@@ -105,7 +107,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.123499e+04   	6.785566e+01   	1.19090
 #####
 # Model for cuda1_impl0 (Comb2)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -117,6 +119,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.123077e+04   	9.504466e+01   	1.179231e+06   	1.324463e+10   	105
 c00cf6b7	29491200       	7.077888e+09   	1.123077e+04   	9.504466e+01   	1.179231e+06   	1.324463e+10   	105
 78a2cc08	29491200       	7.077888e+09   	1.123077e+04   	9.504466e+01   	1.179231e+06   	1.324463e+10   	105
+a7cdf15b	44236800       	1.415578e+10   	2.246154e+04   	1.900893e+02   	2.358462e+06   	5.297852e+10   	105
 24c84a50	11059200       	1.769472e+09   	2.807693e+03   	2.376116e+01   	2.948078e+05   	8.277894e+08   	105
 0b0b0ce8	3686400        	2.621440e+08   	6.672056e+02   	3.376608e+01   	6.805497e+04   	4.552295e+07   	102
 4220e23d	14745600       	2.097152e+09   	5.553764e+03   	3.500896e+02   	5.831453e+05   	3.251521e+09   	105
@@ -143,7 +146,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.123077e+04   	9.504466e+01   	1.17923
 #####
 # Model for cuda2_impl0 (Comb1)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -155,6 +158,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.124174e+04   	2.629960e+01   	1.180383e+06   	1.326963e+10   	105
 c00cf6b7	29491200       	7.077888e+09   	1.124174e+04   	2.629960e+01   	1.180383e+06   	1.326963e+10   	105
 78a2cc08	29491200       	7.077888e+09   	1.124174e+04   	2.629960e+01   	1.180383e+06   	1.326963e+10   	105
+a7cdf15b	44236800       	1.415578e+10   	2.248348e+04   	5.259920e+01   	2.360766e+06   	5.307852e+10   	105
 24c84a50	11059200       	1.769472e+09   	2.810435e+03   	6.574900e+00   	2.950958e+05   	8.293519e+08   	105
 0b0b0ce8	3686400        	2.621440e+08   	6.002221e+02   	2.259043e+01   	6.242310e+04   	3.752080e+07   	104
 4220e23d	14745600       	2.097152e+09   	5.577722e+03   	1.615194e+02   	5.912385e+05   	3.300529e+09   	106

+ 18 - 9
tools/perfmodels/sampling/codelets/45/starpu_sgemm_gemm.idgraf

@@ -28,7 +28,7 @@
 #####
 # Model for cuda0_impl0 (Comb2)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -42,6 +42,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.150361e+04   	5.884814e+02   	1.000814e+06   	1.154310e+10   	87
 c00cf6b7	29491200       	7.077888e+09   	1.150361e+04   	5.884814e+02   	1.000814e+06   	1.154310e+10   	87
 78a2cc08	29491200       	7.077888e+09   	1.150361e+04   	5.884814e+02   	1.000814e+06   	1.154310e+10   	87
+a7cdf15b	44236800       	1.415578e+10   	2.300722e+04   	1.176963e+03   	2.001628e+06   	4.617240e+10   	87
 24c84a50	11059200       	1.769472e+09   	2.875903e+03   	1.471204e+02   	2.502035e+05   	7.214438e+08   	87
 
 ####################
@@ -66,7 +67,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.150361e+04   	5.884814e+02   	1.00081
 #####
 # Model for cuda1_impl0 (Comb4)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -80,6 +81,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.157020e+04   	6.521027e+02   	1.018178e+06   	1.181795e+10   	88
 c00cf6b7	29491200       	7.077888e+09   	1.157020e+04   	6.521027e+02   	1.018178e+06   	1.181795e+10   	88
 78a2cc08	29491200       	7.077888e+09   	1.157020e+04   	6.521027e+02   	1.018178e+06   	1.181795e+10   	88
+a7cdf15b	44236800       	1.415578e+10   	2.314040e+04   	1.304205e+03   	2.036356e+06   	4.727180e+10   	88
 24c84a50	11059200       	1.769472e+09   	2.892550e+03   	1.630257e+02   	2.545445e+05   	7.386219e+08   	88
 
 ####################
@@ -104,7 +106,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.157020e+04   	6.521027e+02   	1.01817
 #####
 # Model for cuda2_impl0 (Comb6)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -118,6 +120,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.162826e+04   	6.757302e+02   	1.023286e+06   	1.193922e+10   	88
 c00cf6b7	29491200       	7.077888e+09   	1.162826e+04   	6.757302e+02   	1.023286e+06   	1.193922e+10   	88
 78a2cc08	29491200       	7.077888e+09   	1.162826e+04   	6.757302e+02   	1.023286e+06   	1.193922e+10   	88
+a7cdf15b	44236800       	1.415578e+10   	2.325652e+04   	1.351460e+03   	2.046572e+06   	4.775688e+10   	88
 24c84a50	11059200       	1.769472e+09   	2.907065e+03   	1.689325e+02   	2.558215e+05   	7.462012e+08   	88
 
 ####################
@@ -142,7 +145,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.162826e+04   	6.757302e+02   	1.02328
 #####
 # Model for cuda3_impl0 (Comb7)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -156,6 +159,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.150498e+04   	4.254093e+02   	9.894285e+05   	1.139892e+10   	86
 c00cf6b7	29491200       	7.077888e+09   	1.150498e+04   	4.254093e+02   	9.894285e+05   	1.139892e+10   	86
 78a2cc08	29491200       	7.077888e+09   	1.150498e+04   	4.254093e+02   	9.894285e+05   	1.139892e+10   	86
+a7cdf15b	44236800       	1.415578e+10   	2.300996e+04   	8.508186e+02   	1.978857e+06   	4.559568e+10   	86
 24c84a50	11059200       	1.769472e+09   	2.876245e+03   	1.063523e+02   	2.473571e+05   	7.124325e+08   	86
 
 ####################
@@ -180,7 +184,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.150498e+04   	4.254093e+02   	9.89428
 #####
 # Model for cuda4_impl0 (Comb0)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -194,6 +198,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.146770e+04   	1.768909e+02   	1.100899e+06   	1.262778e+10   	96
 c00cf6b7	29491200       	7.077888e+09   	1.146770e+04   	1.768909e+02   	1.100899e+06   	1.262778e+10   	96
 78a2cc08	29491200       	7.077888e+09   	1.146770e+04   	1.768909e+02   	1.100899e+06   	1.262778e+10   	96
+a7cdf15b	44236800       	1.415578e+10   	2.293540e+04   	3.537818e+02   	2.201798e+06   	5.051112e+10   	96
 24c84a50	11059200       	1.769472e+09   	2.866925e+03   	4.422272e+01   	2.752248e+05   	7.892362e+08   	96
 
 ####################
@@ -218,7 +223,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.146770e+04   	1.768909e+02   	1.10089
 #####
 # Model for cuda5_impl0 (Comb1)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -232,6 +237,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.149102e+04   	5.375188e+02   	1.114629e+06   	1.283625e+10   	97
 c00cf6b7	29491200       	7.077888e+09   	1.149102e+04   	5.375188e+02   	1.114629e+06   	1.283625e+10   	97
 78a2cc08	29491200       	7.077888e+09   	1.149102e+04   	5.375188e+02   	1.114629e+06   	1.283625e+10   	97
+a7cdf15b	44236800       	1.415578e+10   	2.298204e+04   	1.075038e+03   	2.229258e+06   	5.134500e+10   	97
 24c84a50	11059200       	1.769472e+09   	2.872755e+03   	1.343797e+02   	2.786572e+05   	8.022656e+08   	97
 
 ####################
@@ -256,7 +262,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.149102e+04   	5.375188e+02   	1.11462
 #####
 # Model for cuda6_impl0 (Comb3)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -270,6 +276,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.155069e+04   	5.660846e+02   	1.108866e+06   	1.283893e+10   	96
 c00cf6b7	29491200       	7.077888e+09   	1.155069e+04   	5.660846e+02   	1.108866e+06   	1.283893e+10   	96
 78a2cc08	29491200       	7.077888e+09   	1.155069e+04   	5.660846e+02   	1.108866e+06   	1.283893e+10   	96
+a7cdf15b	44236800       	1.415578e+10   	2.310138e+04   	1.132169e+03   	2.217732e+06   	5.135572e+10   	96
 24c84a50	11059200       	1.769472e+09   	2.887673e+03   	1.415212e+02   	2.772165e+05   	8.024331e+08   	96
 
 ####################
@@ -294,7 +301,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.155069e+04   	5.660846e+02   	1.10886
 #####
 # Model for cuda7_impl0 (Comb5)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -308,6 +315,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.144012e+04   	2.531108e+02   	1.109691e+06   	1.270122e+10   	97
 c00cf6b7	29491200       	7.077888e+09   	1.144012e+04   	2.531108e+02   	1.109691e+06   	1.270122e+10   	97
 78a2cc08	29491200       	7.077888e+09   	1.144012e+04   	2.531108e+02   	1.109691e+06   	1.270122e+10   	97
+a7cdf15b	44236800       	1.415578e+10   	2.288024e+04   	5.062216e+02   	2.219382e+06   	5.080488e+10   	97
 24c84a50	11059200       	1.769472e+09   	2.860030e+03   	6.327770e+01   	2.774228e+05   	7.938262e+08   	97
 
 ####################
@@ -332,7 +340,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.144012e+04   	2.531108e+02   	1.10969
 #####
 # Model for cpu0_impl0 (Comb8)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -346,5 +354,6 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	3.621356e+05   	7.764608e+03   	8.329119e+06   	3.017657e+12   	23
 c00cf6b7	29491200       	7.077888e+09   	3.621356e+05   	7.764608e+03   	8.329119e+06   	3.017657e+12   	23
 78a2cc08	29491200       	7.077888e+09   	3.621356e+05   	7.764608e+03   	8.329119e+06   	3.017657e+12   	23
+a7cdf15b	44236800       	1.415578e+10   	7.242712e+05   	1.552922e+04   	1.665824e+07   	1.207063e+13   	23
 24c84a50	11059200       	1.769472e+09   	9.053390e+04   	1.941152e+03   	2.082280e+06   	1.886036e+11   	23
 

+ 8 - 4
tools/perfmodels/sampling/codelets/45/starpu_sgemm_gemm.mirage

@@ -29,7 +29,7 @@
 #####
 # Model for cpu0_impl0 (Comb3)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -42,6 +42,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	3.550396e+05   	8.949994e+03   	2.840317e+07   	1.009066e+13   	80
 c00cf6b7	29491200       	7.077888e+09   	3.550396e+05   	8.949994e+03   	2.840317e+07   	1.009066e+13   	80
 78a2cc08	29491200       	7.077888e+09   	3.550396e+05   	8.949994e+03   	2.840317e+07   	1.009066e+13   	80
+a7cdf15b	44236800       	1.415578e+10   	7.100792e+05   	1.789999e+04   	5.680634e+07   	4.036264e+13   	80
 24c84a50	11059200       	1.769472e+09   	8.875990e+04   	2.237499e+03   	7.100792e+06   	6.306662e+11   	80
 4220e23d	14745600       	2.097152e+09   	1.078112e+05   	1.983800e+03   	8.624897e+06   	9.301755e+11   	80
 
@@ -67,7 +68,7 @@ c00cf6b7	29491200       	7.077888e+09   	3.550396e+05   	8.949994e+03   	2.84031
 #####
 # Model for cuda0_impl0 (Comb1)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -80,6 +81,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.151398e+04   	9.050114e+01   	1.220482e+06   	1.405348e+10   	106
 c00cf6b7	29491200       	7.077888e+09   	1.151398e+04   	9.050114e+01   	1.220482e+06   	1.405348e+10   	106
 78a2cc08	29491200       	7.077888e+09   	1.151398e+04   	9.050114e+01   	1.220482e+06   	1.405348e+10   	106
+a7cdf15b	44236800       	1.415578e+10   	2.302796e+04   	1.810023e+02   	2.440964e+06   	5.621392e+10   	106
 24c84a50	11059200       	1.769472e+09   	2.878495e+03   	2.262529e+01   	3.051205e+05   	8.783425e+08   	106
 4220e23d	14745600       	2.097152e+09   	5.574713e+03   	3.353004e+02   	5.909196e+05   	3.306125e+09   	106
 
@@ -105,7 +107,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.151398e+04   	9.050114e+01   	1.22048
 #####
 # Model for cuda1_impl0 (Comb0)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -118,6 +120,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.150036e+04   	8.404527e+01   	1.207538e+06   	1.388786e+10   	105
 c00cf6b7	29491200       	7.077888e+09   	1.150036e+04   	8.404527e+01   	1.207538e+06   	1.388786e+10   	105
 78a2cc08	29491200       	7.077888e+09   	1.150036e+04   	8.404527e+01   	1.207538e+06   	1.388786e+10   	105
+a7cdf15b	44236800       	1.415578e+10   	2.300072e+04   	1.680905e+02   	2.415076e+06   	5.555144e+10   	105
 24c84a50	11059200       	1.769472e+09   	2.875090e+03   	2.101132e+01   	3.018845e+05   	8.679912e+08   	105
 4220e23d	14745600       	2.097152e+09   	5.579034e+03   	3.672012e+02   	5.857985e+05   	3.282348e+09   	105
 
@@ -143,7 +146,7 @@ c00cf6b7	29491200       	7.077888e+09   	1.150036e+04   	8.404527e+01   	1.20753
 #####
 # Model for cuda2_impl0 (Comb2)
 # number of entries
-7
+8
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -156,6 +159,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.148096e+04   	7.289415e+01   	1.205501e+06   	1.384086e+10   	105
 c00cf6b7	29491200       	7.077888e+09   	1.148096e+04   	7.289415e+01   	1.205501e+06   	1.384086e+10   	105
 78a2cc08	29491200       	7.077888e+09   	1.148096e+04   	7.289415e+01   	1.205501e+06   	1.384086e+10   	105
+a7cdf15b	44236800       	1.415578e+10   	2.296192e+04   	1.457883e+02   	2.411002e+06   	5.536344e+10   	105
 24c84a50	11059200       	1.769472e+09   	2.870240e+03   	1.822354e+01   	3.013752e+05   	8.650538e+08   	105
 4220e23d	14745600       	2.097152e+09   	5.580581e+03   	3.970717e+02   	5.859610e+05   	3.286558e+09   	105
 

+ 10 - 5
tools/perfmodels/sampling/codelets/45/starpu_sgemm_gemm.sirocco

@@ -28,7 +28,7 @@
 #####
 # Model for cuda0_impl0 (Comb2)
 # number of entries
-8
+9
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -40,6 +40,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	2.745578e+03   	3.064191e+02   	6.616844e+05   	1.839335e+09   	241
 c00cf6b7	29491200       	7.077888e+09   	2.745578e+03   	3.064191e+02   	6.616844e+05   	1.839335e+09   	241
 78a2cc08	29491200       	7.077888e+09   	2.745578e+03   	3.064191e+02   	6.616844e+05   	1.839335e+09   	241
+a7cdf15b	44236800       	1.415578e+10   	5.491156e+03   	6.128382e+02   	1.323369e+06   	7.357340e+09   	241
 24c84a50	11059200       	1.769472e+09   	6.863945e+02   	7.660478e+01   	1.654211e+05   	1.149584e+08   	241
 0b0b0ce8	3686400        	2.621440e+08   	1.582927e+02   	3.333442e+01   	3.434951e+04   	5.678402e+06   	217
 4220e23d	14745600       	2.097152e+09   	8.206871e+02   	1.017181e+02   	1.148962e+05   	9.574235e+07   	140
@@ -67,7 +68,7 @@ c00cf6b7	29491200       	7.077888e+09   	2.745578e+03   	3.064191e+02   	6.61684
 #####
 # Model for cuda3_impl0 (Comb1)
 # number of entries
-8
+9
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -79,6 +80,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	2.686428e+03   	2.002215e+02   	6.716071e+05   	1.814247e+09   	250
 c00cf6b7	29491200       	7.077888e+09   	2.686428e+03   	2.002215e+02   	6.716071e+05   	1.814247e+09   	250
 78a2cc08	29491200       	7.077888e+09   	2.686428e+03   	2.002215e+02   	6.716071e+05   	1.814247e+09   	250
+a7cdf15b	44236800       	1.415578e+10   	5.372856e+03   	4.004430e+02   	1.343214e+06   	7.256988e+09   	251
 24c84a50	11059200       	1.769472e+09   	6.716070e+02   	5.005537e+01   	1.679018e+05   	1.133904e+08   	250
 0b0b0ce8	3686400        	2.621440e+08   	1.630480e+02   	3.438768e+01   	3.097912e+04   	5.275762e+06   	190
 4220e23d	14745600       	2.097152e+09   	8.448030e+02   	7.773742e+01   	2.433033e+05   	2.072837e+08   	288
@@ -106,7 +108,7 @@ c00cf6b7	29491200       	7.077888e+09   	2.686428e+03   	2.002215e+02   	6.71607
 #####
 # Model for cuda1_impl0 (Comb0)
 # number of entries
-8
+9
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -118,6 +120,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	2.791098e+03   	3.147711e+02   	6.503258e+05   	1.838209e+09   	233
 c00cf6b7	29491200       	7.077888e+09   	2.791098e+03   	3.147711e+02   	6.503258e+05   	1.838209e+09   	233
 78a2cc08	29491200       	7.077888e+09   	2.791098e+03   	3.147711e+02   	6.503258e+05   	1.838209e+09   	233
+a7cdf15b	44236800       	1.415578e+10   	5.582196e+03   	6.295422e+02   	1.300652e+06   	7.352836e+09   	233
 24c84a50	11059200       	1.769472e+09   	6.977745e+02   	7.869277e+01   	1.625815e+05   	1.148881e+08   	233
 0b0b0ce8	3686400        	2.621440e+08   	1.624855e+02   	3.298013e+01   	2.940987e+04   	4.975550e+06   	181
 4220e23d	14745600       	2.097152e+09   	8.152506e+02   	1.017614e+02   	1.173961e+05   	9.719839e+07   	144
@@ -145,7 +148,7 @@ c00cf6b7	29491200       	7.077888e+09   	2.791098e+03   	3.147711e+02   	6.50325
 #####
 # Model for cuda2_impl0 (Comb3)
 # number of entries
-8
+9
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -157,6 +160,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	2.754203e+03   	2.682327e+02   	6.830422e+05   	1.899080e+09   	248
 c00cf6b7	29491200       	7.077888e+09   	2.754203e+03   	2.682327e+02   	6.830422e+05   	1.899080e+09   	248
 78a2cc08	29491200       	7.077888e+09   	2.754203e+03   	2.682327e+02   	6.830422e+05   	1.899080e+09   	248
+a7cdf15b	44236800       	1.415578e+10   	5.508406e+03   	5.364654e+02   	1.366084e+06   	7.596320e+09   	248
 24c84a50	11059200       	1.769472e+09   	6.885507e+02   	6.705818e+01   	1.707605e+05   	1.186925e+08   	248
 0b0b0ce8	3686400        	2.621440e+08   	1.622246e+02   	3.553894e+01   	3.714942e+04   	6.315779e+06   	229
 4220e23d	14745600       	2.097152e+09   	8.611626e+02   	9.290485e+01   	2.411255e+05   	2.100651e+08   	280
@@ -184,7 +188,7 @@ c00cf6b7	29491200       	7.077888e+09   	2.754203e+03   	2.682327e+02   	6.83042
 #####
 # Model for cpu0_impl0 (Comb4)
 # number of entries
-8
+9
 # sumlnx	sumlnx2		sumlny		sumlnxlny	alpha		beta		n	minx		maxx
 0.000000e+00   	0.000000e+00   	0.000000e+00   	0.000000e+00   	nan            	nan            	0	0              	0              
 # a		b		c
@@ -196,6 +200,7 @@ nan            	nan            	nan
 9c6670ef	29491200       	7.077888e+09   	1.712078e+05   	4.163047e+04   	2.773567e+07   	5.029326e+12   	162
 c00cf6b7	29491200       	7.077888e+09   	1.712078e+05   	4.163047e+04   	2.773567e+07   	5.029326e+12   	162
 78a2cc08	29491200       	7.077888e+09   	1.712078e+05   	4.163047e+04   	2.773567e+07   	5.029326e+12   	162
+a7cdf15b	44236800       	1.415578e+10   	3.424156e+05   	8.326094e+04   	5.547134e+07   	2.011730e+13   	162
 24c84a50	11059200       	1.769472e+09   	4.280195e+04   	1.040762e+04   	6.933918e+06   	3.143329e+11   	162
 0b0b0ce8	3686400        	2.621440e+08   	6.441655e+03   	1.152866e+03   	3.220827e+05   	2.141201e+09   	50
 4220e23d	14745600       	2.097152e+09   	4.927734e+04   	1.166029e+04   	5.913281e+06   	3.077063e+11   	120