[Math-atlas-commits] CVS: AtlasBase/Clint/misc/ARCHDEF/Corei764SSE3/gemm/gcc cCleanK, NONE, 1.1 cC
Brought to you by:
rwhaley,
tonyc040457
Update of /cvsroot/math-atlas/AtlasBase/Clint/misc/ARCHDEF/Corei764SSE3/gemm/gcc In directory 23jxhf1.ch3.sourceforge.com:/tmp/cvs-serv14578/Clint/misc/ARCHDEF/Corei764SSE3/gemm/gcc Added Files: cCleanK cCleanM cCleanN cMMRES cNB cNCNB cXover.h cbestNN_0x0x0 cbestNN_0x0x40 cbestNN_40x40x40 cbestNT_0x0x0 cbestNT_0x0x40 cbestNT_40x40x40 cbestTN_0x0x0 cbestTN_0x0x40 cbestTN_40x40x40 cbestTT_0x0x0 cbestTT_0x0x40 cbestTT_40x40x40 cgMMRES cuCleanK cuCleanKF cuCleanM cuCleanMF cuCleanN cuCleanNF cuMMRES dCleanK dCleanM dCleanN dMMRES dNB dNCNB dXover.h dbestNN_0x0x0 dbestNN_0x0x44 dbestNN_44x44x44 dbestNT_0x0x0 dbestNT_0x0x44 dbestNT_44x44x44 dbestTN_0x0x0 dbestTN_0x0x44 dbestTN_44x44x44 dbestTT_0x0x0 dbestTT_0x0x44 dbestTT_44x44x44 dgMMRES dnreg duCleanK duCleanKF duCleanM duCleanMF duCleanN duCleanNF duMMRES sCleanK sCleanM sCleanN sMMRES sNB sNCNB sXover.h sbestNN_0x0x0 sbestNN_0x0x48 sbestNN_48x48x48 sbestNT_0x0x0 sbestNT_0x0x48 sbestNT_48x48x48 sbestTN_0x0x0 sbestTN_0x0x48 sbestTN_48x48x48 sbestTT_0x0x0 sbestTT_0x0x48 sbestTT_48x48x48 sgMMRES snreg suCleanK suCleanKF suCleanM suCleanMF suCleanN suCleanNF suMMRES zCleanK zCleanM zCleanN zMMRES zNB zNCNB zXover.h zbestNN_0x0x0 zbestNN_0x0x32 zbestNN_32x32x32 zbestNT_0x0x0 zbestNT_0x0x32 zbestNT_32x32x32 zbestTN_0x0x0 zbestTN_0x0x32 zbestTN_32x32x32 zbestTT_0x0x0 zbestTT_0x0x32 zbestTT_32x32x32 zgMMRES zuCleanK zuCleanKF zuCleanM zuCleanMF zuCleanN zuCleanNF zuMMRES Log Message: --- NEW FILE: cCleanK --- KB MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH GEN-MFLOP SPC-MFLOP 80 1 4 512 80 4 2 16 0 4 2 -4943.91 -4542.59 79 1 4 512 80 4 2 16 0 4 2 -4879.98 -4542.59 78 1 4 512 80 4 2 16 0 4 2 -4880.28 -4542.59 77 1 4 512 80 4 2 16 0 4 2 -4896.03 -4542.59 76 1 4 512 80 4 2 16 0 4 2 -4875.95 -4542.59 75 1 4 512 80 4 2 16 0 4 2 -4883.07 -4542.59 74 1 4 512 80 4 2 16 0 4 2 -4881.93 -4542.59 73 1 4 512 80 4 2 16 0 4 2 -4865.29 -4542.59 72 1 4 512 80 4 2 16 0 4 2 -4875.87 -4542.59 71 1 4 512 80 4 2 16 0 4 2 -4852.30 -4542.59 70 1 4 512 80 4 2 16 0 4 2 -4858.29 -4542.59 69 1 4 512 80 4 2 16 0 4 2 -4870.96 -4542.59 68 1 4 512 80 4 2 16 0 4 2 -4859.07 -4542.59 67 1 4 512 80 4 2 16 0 4 2 -4824.00 -4542.59 66 1 4 512 80 4 2 16 0 4 2 -4774.74 -4542.59 65 1 4 512 80 4 2 16 0 4 2 -4864.50 -4542.59 64 1 4 512 80 4 2 16 0 4 2 -4891.00 -4542.59 63 1 4 512 80 4 2 16 0 4 2 -4821.74 -4542.59 62 1 4 512 80 4 2 16 0 4 2 -4827.03 -4542.59 61 1 4 512 80 4 2 16 0 4 2 -4831.71 -4542.59 60 1 4 512 80 4 2 16 0 4 2 -4825.38 -4542.59 59 1 4 512 80 4 2 16 0 4 2 -4814.01 -4542.59 58 1 4 512 80 4 2 16 0 4 2 -4795.57 -4542.59 57 1 4 512 80 4 2 16 0 4 2 -4811.26 -4542.59 56 1 4 512 80 4 2 16 0 4 2 -4788.31 -4542.59 55 1 4 512 80 4 2 16 0 4 2 -4772.97 -4542.59 54 1 4 512 80 4 2 16 0 4 2 -4786.83 -4542.59 53 1 4 512 80 4 2 16 0 4 2 -4798.92 -4542.59 52 1 4 512 80 4 2 16 0 4 2 -4772.88 -4542.59 51 1 4 512 80 4 2 16 0 4 2 -4740.38 -4542.59 50 1 4 512 80 4 2 16 0 4 2 -4714.89 -4542.59 49 1 4 512 80 4 2 16 0 4 2 -4788.24 -4542.59 48 1 4 512 80 4 2 16 0 4 2 -4855.54 -4542.59 47 1 4 512 80 4 2 16 0 4 2 -4748.91 -4542.59 46 1 4 512 80 4 2 16 0 4 2 -4745.42 -4542.59 45 1 4 512 80 4 2 16 0 4 2 -4763.26 -4542.59 44 1 4 512 80 4 2 16 0 4 2 -4687.12 -4542.59 43 1 4 512 80 4 2 16 0 4 2 -4714.13 -4542.59 42 1 4 512 80 4 2 16 0 4 2 -4707.28 -4542.59 41 1 4 512 80 4 2 16 0 4 2 -4675.51 -4542.59 40 1 4 512 80 4 2 16 0 4 2 -4673.56 -4542.59 39 1 4 512 80 4 2 16 0 4 2 -4696.69 -4542.59 38 1 4 512 80 4 2 16 0 4 2 -4659.50 -4542.59 37 1 4 512 80 4 2 16 0 4 2 -4655.11 -4542.59 36 1 4 512 80 4 2 16 0 4 2 -4662.26 -4542.59 35 1 4 512 80 4 2 16 0 4 2 -4657.92 -4542.59 34 1 4 512 80 4 2 16 0 4 2 -4599.96 -4542.59 33 1 4 512 80 4 2 16 0 4 2 -4672.64 -4542.59 32 1 4 512 80 4 2 16 0 4 2 -4752.39 -4542.59 31 1 4 512 80 4 2 16 0 4 2 -4808.33 -4542.59 30 1 4 512 80 4 2 16 0 4 2 -4808.48 -4542.59 29 1 4 512 80 4 2 16 0 4 2 -4795.30 -4542.59 28 1 4 512 80 4 2 16 0 4 2 -4795.09 -4542.59 27 1 4 512 80 4 2 16 0 4 2 -4838.18 -4542.59 26 1 4 512 80 4 2 16 0 4 2 -4762.04 -4542.59 25 1 4 512 80 4 2 16 0 4 2 -4737.60 -4542.59 24 1 4 512 80 4 2 16 0 4 2 -4713.73 -4542.59 23 1 4 512 80 4 2 16 0 4 2 -4755.69 -4542.59 22 1 4 512 80 4 2 16 0 4 2 -4689.20 -4542.59 21 1 4 512 80 4 2 16 0 4 2 -4669.14 -4542.59 20 1 4 512 80 4 2 16 0 4 2 -4678.54 -4542.59 19 1 4 512 80 4 2 16 0 4 2 -4733.45 -4542.59 18 1 4 512 80 4 2 16 0 4 2 -4631.07 -4542.59 17 1 4 512 80 4 2 16 0 4 2 -4615.51 -4542.59 0 1 4 512 80 4 2 12 0 4 2 -4560.50 -4542.59 --- NEW FILE: cCleanM --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 80 4 2 16 0 4 2 -4912.45 --- NEW FILE: cCleanN --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 80 4 2 16 0 4 2 -4934.89 --- NEW FILE: cMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 80 4 2 16 0 4 2 -4980.53 ICASE NB MFLOP ROUT AUTHOR 223 80 -15609.62 "ATL_smm2x2x256_sse.c" "R. Clint Whaley" --- NEW FILE: cNB --- 1 80 --- NEW FILE: cNCNB --- 40 --- NEW FILE: cXover.h --- #ifndef CXOVER_H #define CXOVER_H #define ATL_3NB 240 #define NN_MNK_M 18000 #define NN_MNK_N 18000 #define NN_MNK_MN 96000 #define NN_MNK_K 46080 #define NN_MNK_GE 3375 #define NT_MNK_M 18000 #define NT_MNK_N 18000 #define NT_MNK_MN 64000 #define NT_MNK_K 18000 #define NT_MNK_GE 3375 #define TN_MNK_M 18000 #define TN_MNK_N 8000 #define TN_MNK_MN 64000 #define TN_MNK_K 8000 #define TN_MNK_GE 3375 #define TT_MNK_M 18000 #define TT_MNK_N 18000 #define TT_MNK_MN 64000 #define TT_MNK_K 18000 #define TT_MNK_GE 3375 #define C2R_K 284 #endif --- NEW FILE: cbestNN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 40 4 2 1 0 4 2 -4344.89 --- NEW FILE: cbestNN_0x0x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 512 40 4 2 40 0 4 2 -4651.75 --- NEW FILE: cbestNN_40x40x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 40 4 2 40 0 4 2 -4662.33 --- NEW FILE: cbestNT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 40 4 2 12 0 4 2 -4000.81 --- NEW FILE: cbestNT_0x0x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 512 40 4 2 40 0 4 2 -4379.39 --- NEW FILE: cbestNT_40x40x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 40 4 2 40 0 4 2 -4392.86 --- NEW FILE: cbestTN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 512 40 4 2 12 0 4 2 -4209.39 --- NEW FILE: cbestTN_0x0x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 40 4 2 40 0 4 2 -4511.09 --- NEW FILE: cbestTN_40x40x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 40 4 2 40 0 4 2 -4668.00 --- NEW FILE: cbestTT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 40 4 2 12 0 4 2 -4099.66 --- NEW FILE: cbestTT_0x0x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 40 4 2 40 0 4 2 -4512.40 --- NEW FILE: cbestTT_40x40x40 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 40 4 2 40 0 4 2 -4563.67 --- NEW FILE: cgMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 72 4 2 72 0 8 1 -4954.54 --- NEW FILE: cuCleanK --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 1 0 80 70 40 16 -3004.65 ATL_mm4x4x2US.c 4 221 1 80 68 40 16 -13400.57 ATL_smm10x1x120_sse.c --- NEW FILE: cuCleanKF --- MULT ICASE FIXED NB 1 4 221 1 80 --- NEW FILE: cuCleanM --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 326 0 80 70 40 16 -2149.85 ATL_mm6x8x8_1p.c 2 223 0 80 70 40 16 -15833.27 ATL_smm2x2x256_sse.c --- NEW FILE: cuCleanMF --- MULT ICASE FIXED NB 1 2 223 0 80 --- NEW FILE: cuCleanN --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 222 0 80 70 40 16 -14541.33 ATL_smm4x1x256_sse.c 2 223 0 80 70 40 16 -15622.35 ATL_smm2x2x256_sse.c --- NEW FILE: cuCleanNF --- MULT ICASE FIXED NB 2 1 222 0 80 2 223 0 80 --- NEW FILE: cuMMRES --- CASE NB MFLOP ROUTINE 223 80 -15609.62 "ATL_smm2x2x256_sse.c" "R. Clint Whaley" --- NEW FILE: dCleanK --- KB MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH GEN-MFLOP SPC-MFLOP 60 1 5 513 60 4 2 60 1 8 1 -4789.02 -4101.29 59 1 5 513 60 4 2 59 1 8 1 -4736.67 -4101.29 58 1 5 513 60 4 2 58 1 8 1 -4777.20 -4101.29 57 1 5 513 60 4 2 57 1 8 1 -4762.95 -4101.29 56 1 5 513 60 4 2 56 1 8 1 -4770.17 -4101.29 55 1 5 513 60 4 2 55 1 8 1 -4759.54 -4101.29 54 1 5 513 60 4 2 54 1 8 1 -4792.51 -4101.29 53 1 5 513 60 4 2 53 1 8 1 -4737.81 -4101.29 52 1 5 513 60 4 2 52 1 8 1 -4777.96 -4101.29 51 1 5 513 60 4 2 51 1 8 1 -4759.08 -4101.29 50 1 5 513 60 4 2 50 1 8 1 -4769.18 -4101.29 49 1 5 513 60 4 2 49 1 8 1 -4770.62 -4101.29 48 1 5 513 60 4 2 48 1 8 1 -4761.88 -4101.29 47 1 5 513 60 4 2 47 1 8 1 -4762.78 -4101.29 46 1 5 513 60 4 2 46 1 8 1 -4779.64 -4101.29 45 1 5 513 60 4 2 45 1 8 1 -4738.58 -4101.29 44 1 5 513 60 4 2 44 1 8 1 -4765.66 -4101.29 43 1 5 513 60 4 2 43 1 8 1 -4758.35 -4101.29 42 1 5 513 60 4 2 42 1 8 1 -4772.65 -4101.29 41 1 5 513 60 4 2 41 1 8 1 -4743.46 -4101.29 40 1 5 513 60 4 2 40 1 8 1 -4740.98 -4101.29 39 1 5 513 60 4 2 39 1 8 1 -4731.23 -4101.29 38 1 5 513 60 4 2 38 1 8 1 -4773.81 -4101.29 37 1 5 513 60 4 2 37 1 8 1 -4729.25 -4101.29 36 1 5 513 60 4 2 36 1 8 1 -4725.26 -4101.29 35 1 5 513 60 4 2 35 1 8 1 -4717.15 -4101.29 34 1 5 513 60 4 2 34 1 8 1 -4752.71 -4101.29 33 1 5 513 60 4 2 33 1 8 1 -4704.96 -4101.29 32 1 5 513 60 4 2 32 1 8 1 -4735.06 -4101.29 31 1 5 513 60 4 2 31 1 8 1 -4683.55 -4101.29 30 1 5 513 60 4 2 30 1 8 1 -4750.99 -4101.29 29 1 5 513 60 4 2 29 1 8 1 -4677.52 -4101.29 28 1 5 513 60 4 2 28 1 8 1 -4735.40 -4101.29 27 1 5 513 60 4 2 27 1 8 1 -4666.62 -4101.29 26 1 5 513 60 4 2 26 1 8 1 -4706.44 -4101.29 25 1 5 513 60 4 2 25 1 8 1 -4681.02 -4101.29 24 1 5 513 60 4 2 24 1 8 1 -4677.55 -4101.29 23 1 5 513 60 4 2 23 1 8 1 -4677.01 -4101.29 22 1 5 513 60 4 2 22 1 8 1 -4720.77 -4101.29 21 1 5 513 60 4 2 21 1 8 1 -4625.61 -4101.29 20 1 5 513 60 4 2 20 1 8 1 -4648.31 -4101.29 19 1 5 513 60 4 2 19 1 8 1 -4624.82 -4101.29 18 1 5 513 60 4 2 18 1 8 1 -4643.89 -4101.29 17 1 5 513 60 4 2 17 1 8 1 -4590.08 -4101.29 16 1 5 513 60 4 2 16 1 8 1 -4610.96 -4101.29 15 1 5 513 60 4 2 15 1 8 1 -4543.17 -4101.29 14 1 5 513 60 4 2 14 1 8 1 -4594.46 -4101.29 13 1 5 513 60 4 2 13 1 8 1 -4453.17 -4101.29 12 1 5 513 60 4 2 12 1 8 1 -4479.33 -4101.29 11 1 5 513 60 4 2 11 1 8 1 -4409.06 -4101.29 10 1 5 513 60 4 2 10 1 8 1 -4412.00 -4101.29 9 1 5 513 60 4 2 9 1 8 1 -4280.61 -4101.29 8 1 5 513 60 4 2 8 1 8 1 -4323.29 -4101.29 0 1 5 513 60 4 2 1 1 8 1 -4107.74 -4101.29 --- NEW FILE: dCleanM --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 60 4 2 60 1 8 1 -4730.69 --- NEW FILE: dCleanN --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 60 4 2 60 1 8 1 -4733.48 --- NEW FILE: dMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 60 4 2 60 1 8 1 -4730.15 ICASE NB MFLOP ROUT AUTHOR 241 60 -9094.79 "ATL_dmm4x2x128_sse2.c" "Whaley & Voronenko" --- NEW FILE: dNB --- 1 60 --- NEW FILE: dNCNB --- 44 --- NEW FILE: dXover.h --- #ifndef DXOVER_H #define DXOVER_H #define ATL_3NB 180 #define NN_MNK_M 13500 #define NN_MNK_N 6000 #define NN_MNK_MN 36000 #define NN_MNK_K 132540 #define NN_MNK_GE 103823 #define NT_MNK_M 6000 #define NT_MNK_N 6000 #define NT_MNK_MN 36000 #define NT_MNK_K 54000 #define NT_MNK_GE 13824 #define TN_MNK_M 6000 #define TN_MNK_N 6000 #define TN_MNK_MN 36000 #define TN_MNK_K 6000 #define TN_MNK_GE 3375 #define TT_MNK_M 6000 #define TT_MNK_N 6000 #define TT_MNK_MN 36000 #define TT_MNK_K 54000 #define TT_MNK_GE 13824 #endif --- NEW FILE: dbestNN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 44 4 2 22 1 8 1 -4307.12 --- NEW FILE: dbestNN_0x0x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 513 44 4 2 44 1 8 1 -4726.36 --- NEW FILE: dbestNN_44x44x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 513 44 4 2 44 1 8 1 -4637.83 --- NEW FILE: dbestNT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 44 4 2 22 1 8 1 -4046.25 --- NEW FILE: dbestNT_0x0x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 513 44 4 2 44 1 8 1 -4326.42 --- NEW FILE: dbestNT_44x44x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 513 44 4 2 44 1 8 1 -4328.51 --- NEW FILE: dbestTN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 513 44 4 2 1 1 8 1 -4066.44 --- NEW FILE: dbestTN_0x0x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 513 44 4 2 44 1 8 1 -4331.46 --- NEW FILE: dbestTN_44x44x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 44 4 2 44 1 8 1 -4296.23 --- NEW FILE: dbestTT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 44 4 2 22 1 8 1 -4234.35 --- NEW FILE: dbestTT_0x0x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 513 44 4 2 44 1 8 1 -4504.50 --- NEW FILE: dbestTT_44x44x44 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 44 4 2 44 1 8 1 -4511.62 --- NEW FILE: dgMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 60 4 2 60 1 8 1 -4730.15 --- NEW FILE: dnreg --- 19 --- NEW FILE: duCleanK --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 327 1 60 53 30 16 -3768.63 ATL_dmm4x1x90_x87.c 2 241 1 60 52 30 16 -8425.33 ATL_dmm4x2x128_sse2.c --- NEW FILE: duCleanKF --- MULT ICASE FIXED NB 1 2 241 1 60 --- NEW FILE: duCleanM --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 2 243 0 60 52 30 16 -8969.47 ATL_dmm2x2x128_sse2.c 4 241 0 60 52 28 16 -9008.71 ATL_dmm4x2x128_sse2.c --- NEW FILE: duCleanMF --- MULT ICASE FIXED NB 2 2 243 0 60 4 241 0 60 --- NEW FILE: duCleanN --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 228 0 60 53 30 16 -8073.46 ATL_dmm6x1x72_sse2.c 2 241 0 60 52 30 16 -8951.57 ATL_dmm4x2x128_sse2.c --- NEW FILE: duCleanNF --- MULT ICASE FIXED NB 2 1 228 0 60 2 241 0 60 --- NEW FILE: duMMRES --- CASE NB MFLOP ROUTINE 241 60 -9094.79 "ATL_dmm4x2x128_sse2.c" "Whaley & Voronenko" --- NEW FILE: sCleanK --- KB MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH GEN-MFLOP SPC-MFLOP 72 1 4 512 72 4 2 72 0 8 1 -4914.64 -4143.83 71 1 4 512 72 4 2 71 0 8 1 -4969.23 -4144.12 70 1 4 512 72 4 2 70 0 8 1 -4971.83 -4132.10 69 1 4 512 72 4 2 69 0 8 1 -4972.14 -4121.99 68 1 4 512 72 4 2 68 0 8 1 -4922.27 -4117.37 67 1 4 512 72 4 2 67 0 8 1 -4971.08 -4110.36 66 1 4 512 72 4 2 66 0 8 1 -4928.14 -4138.38 65 1 4 512 72 4 2 65 0 8 1 -4924.27 -4138.37 64 1 4 512 72 4 2 64 0 8 1 -4931.14 -4121.21 63 1 4 512 72 4 2 63 0 8 1 -4914.85 -4122.77 62 1 4 512 72 4 2 62 0 8 1 -4967.16 -4124.39 61 1 4 512 72 4 2 61 0 8 1 -4974.20 -4103.51 60 1 4 512 72 4 2 60 0 8 1 -4984.91 -4104.27 59 1 4 512 72 4 2 59 0 8 1 -4971.92 -4109.53 58 1 4 512 72 4 2 58 0 8 1 -4978.83 -4101.86 57 1 4 512 72 4 2 57 0 8 1 -4954.35 -4105.08 56 1 4 512 72 4 2 56 0 8 1 -4966.40 -4112.94 55 1 4 512 72 4 2 55 0 8 1 -4965.60 -4084.77 54 1 4 512 72 4 2 54 0 8 1 -4977.50 -4090.04 53 1 4 512 72 4 2 53 0 8 1 -4968.31 -4069.19 52 1 4 512 72 4 2 52 0 8 1 -4956.19 -4069.95 51 1 4 512 72 4 2 51 0 8 1 -4965.94 -4074.80 50 1 4 512 72 4 2 50 0 8 1 -4977.11 -4060.42 49 1 4 512 72 4 2 49 0 8 1 -4969.21 -4058.27 48 1 4 512 72 4 2 48 0 8 1 -4935.85 -4048.02 47 1 4 512 72 4 2 47 0 8 1 -4927.93 -4028.48 46 1 4 512 72 4 2 46 0 8 1 -4976.57 -4006.87 45 1 4 512 72 4 2 45 0 8 1 -4960.78 -4012.11 44 1 4 512 72 4 2 44 0 8 1 -4977.02 -4040.99 43 1 4 512 72 4 2 43 0 8 1 -4959.91 -4040.68 42 1 4 512 72 4 2 42 0 8 1 -4967.42 -4023.88 41 1 4 512 72 4 2 41 0 8 1 -4938.40 -3991.28 40 1 4 512 72 4 2 40 0 8 1 -4938.00 -4009.61 39 1 4 512 72 4 2 39 0 8 1 -4943.62 -3980.54 38 1 4 512 72 4 2 38 0 8 1 -4972.82 -3987.65 37 1 4 512 72 4 2 37 0 8 1 -4918.66 -3979.04 36 1 4 512 72 4 2 36 0 8 1 -4950.21 -3973.77 35 1 4 512 72 4 2 35 0 8 1 -4954.81 -3967.90 34 1 4 512 72 4 2 34 0 8 1 -4961.67 -3971.98 33 1 4 512 72 4 2 33 0 8 1 -4935.43 -3938.25 32 1 4 512 72 4 2 32 0 8 1 -4976.94 -3931.12 31 1 4 512 72 4 2 31 0 8 1 -4870.40 -3924.21 30 1 4 512 72 4 2 30 0 8 1 -4938.84 -3892.03 29 1 4 512 72 4 2 29 0 8 1 -4953.00 -3891.01 28 1 4 512 72 4 2 28 0 8 1 -4928.15 -3872.34 27 1 4 512 72 4 2 27 0 8 1 -4924.84 -3850.15 26 1 4 512 72 4 2 26 0 8 1 -4908.69 -3818.07 25 1 4 512 72 4 2 25 0 8 1 -4890.49 -3790.94 24 1 4 512 72 4 2 24 0 8 1 -4895.50 -3759.16 23 1 4 512 72 4 2 23 0 8 1 -4888.29 -3746.73 22 1 4 512 72 4 2 22 0 8 1 -4870.33 -3773.72 21 1 4 512 72 4 2 21 0 8 1 -4770.72 -3752.20 20 1 4 512 72 4 2 20 0 8 1 -4879.27 -3751.62 19 1 4 512 72 4 2 19 0 8 1 -4771.49 -3746.73 18 1 4 512 72 4 2 18 0 8 1 -4849.86 -3746.73 17 1 4 512 72 4 2 17 0 8 1 -4758.33 -3746.73 16 1 4 512 72 4 2 16 0 8 1 -4896.06 -3746.73 15 1 4 512 72 4 2 15 0 8 1 -4710.23 -3746.73 14 1 4 512 72 4 2 14 0 8 1 -4780.97 -3746.73 13 1 4 512 72 4 2 13 0 8 1 -4580.98 -3746.73 12 1 4 512 72 4 2 12 0 8 1 -4681.18 -3746.73 11 1 4 512 72 4 2 11 0 8 1 -4561.37 -3746.73 10 1 4 512 72 4 2 10 0 8 1 -4579.39 -3746.73 9 1 4 512 72 4 2 9 0 8 1 -4455.30 -3746.73 8 1 4 512 72 4 2 8 0 8 1 -4516.21 -3746.73 7 1 4 512 72 4 2 7 0 8 1 -4296.28 -3746.73 6 1 4 512 72 4 2 6 0 8 1 -4315.55 -3746.73 5 1 4 512 72 4 2 5 0 8 1 -4036.30 -3746.73 4 1 4 512 72 4 2 4 0 8 1 -4062.51 -3746.73 3 1 4 512 72 4 2 3 0 8 1 -3916.62 -3746.73 0 1 4 512 72 4 2 1 0 8 1 -3466.36 -3746.73 --- NEW FILE: sCleanM --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 72 4 2 72 0 8 1 -4968.40 --- NEW FILE: sCleanN --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 72 4 2 72 0 8 1 -4930.17 --- NEW FILE: sMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 72 4 2 72 0 8 1 -4976.39 ICASE NB MFLOP ROUT AUTHOR 223 72 -17000.91 "ATL_smm2x2x256_sse.c" "R. Clint Whaley" --- NEW FILE: sNB --- 1 72 --- NEW FILE: sNCNB --- 48 --- NEW FILE: sXover.h --- #ifndef SXOVER_H #define SXOVER_H #define ATL_3NB 216 #define NN_MNK_M 7200 #define NN_MNK_N 7200 #define NN_MNK_MN 51840 #define NN_MNK_K 64800 #define NN_MNK_GE 13824 #define NT_MNK_M 7200 #define NT_MNK_N 7200 #define NT_MNK_MN 51840 #define NT_MNK_K 41472 #define NT_MNK_GE 13824 #define TN_MNK_M 7200 #define TN_MNK_N 7200 #define TN_MNK_MN 51840 #define TN_MNK_K 16200 #define TN_MNK_GE 13824 #define TT_MNK_M 7200 #define TT_MNK_N 7200 #define TT_MNK_MN 51840 #define TT_MNK_K 41472 #define TT_MNK_GE 13824 #endif --- NEW FILE: sbestNN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 48 4 2 1 0 8 1 -4464.56 --- NEW FILE: sbestNN_0x0x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 3 512 48 4 2 48 0 8 1 -4708.67 --- NEW FILE: sbestNN_48x48x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 512 48 4 2 48 0 8 1 -4713.51 --- NEW FILE: sbestNT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 48 4 2 16 0 8 1 -4197.83 --- NEW FILE: sbestNT_0x0x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 512 48 4 2 48 0 8 1 -4465.70 --- NEW FILE: sbestNT_48x48x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 48 4 2 48 0 8 1 -4460.26 --- NEW FILE: sbestTN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 48 4 2 1 0 8 1 -4098.95 --- NEW FILE: sbestTN_0x0x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 3 512 48 4 2 48 0 8 1 -4245.94 --- NEW FILE: sbestTN_48x48x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 6 512 48 4 2 48 0 8 1 -4340.09 --- NEW FILE: sbestTT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 512 48 4 2 20 0 8 1 -4313.29 --- NEW FILE: sbestTT_0x0x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 512 48 4 2 48 0 8 1 -4632.85 --- NEW FILE: sbestTT_48x48x48 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 512 48 4 2 48 0 8 1 -4593.11 --- NEW FILE: sgMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 512 72 4 2 72 0 8 1 -4976.39 --- NEW FILE: snreg --- 17 --- NEW FILE: suCleanK --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 1 0 72 63 36 16 -3041.12 ATL_mm4x4x2US.c 4 223 1 72 60 36 16 -13892.89 ATL_smm2x2x256_sse.c --- NEW FILE: suCleanKF --- MULT ICASE FIXED NB 1 4 223 1 72 --- NEW FILE: suCleanM --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 326 0 72 63 36 16 -2161.98 ATL_mm6x8x8_1p.c 2 223 0 72 62 36 16 -16997.43 ATL_smm2x2x256_sse.c --- NEW FILE: suCleanMF --- MULT ICASE FIXED NB 1 2 223 0 72 --- NEW FILE: suCleanN --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 220 0 72 63 36 16 -15342.13 ATL_smm6x1x80_sse.c 2 223 0 72 62 36 16 -17005.33 ATL_smm2x2x256_sse.c --- NEW FILE: suCleanNF --- MULT ICASE FIXED NB 2 1 220 0 72 2 223 0 72 --- NEW FILE: suMMRES --- CASE NB MFLOP ROUTINE 223 72 -17000.91 "ATL_smm2x2x256_sse.c" "R. Clint Whaley" --- NEW FILE: zCleanK --- KB MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH GEN-MFLOP SPC-MFLOP 52 1 5 513 52 4 2 52 0 4 2 -4731.27 -4296.13 51 1 5 513 52 4 2 51 0 4 2 -4747.71 -4286.19 50 1 5 513 52 4 2 50 0 4 2 -4739.59 -4292.19 49 1 5 513 52 4 2 49 0 4 2 -4726.15 -4261.31 48 1 5 513 52 4 2 48 0 4 2 -4716.89 -4308.36 47 1 5 513 52 4 2 47 0 4 2 -4734.36 -4276.18 46 1 5 513 52 4 2 46 0 4 2 -4713.72 -4265.55 45 1 5 513 52 4 2 45 0 4 2 -4702.25 -4285.80 44 1 5 513 52 4 2 44 0 4 2 -4691.32 -4241.55 43 1 5 513 52 4 2 43 0 4 2 -4723.45 -4250.41 42 1 5 513 52 4 2 42 0 4 2 -4694.61 -4224.88 41 1 5 513 52 4 2 41 0 4 2 -4682.23 -4205.82 40 1 5 513 52 4 2 40 0 4 2 -4689.89 -4198.74 39 1 5 513 52 4 2 39 0 4 2 -4698.15 -4192.48 38 1 5 513 52 4 2 38 0 4 2 -4699.70 -4164.92 37 1 5 513 52 4 2 37 0 4 2 -4683.38 -4163.04 36 1 5 513 52 4 2 36 0 4 2 -4656.96 -4198.59 35 1 5 513 52 4 2 35 0 4 2 -4654.22 -4078.12 34 1 5 513 52 4 2 34 0 4 2 -4572.88 -4082.41 0 1 5 513 52 4 2 1 0 4 2 -3932.77 -4078.12 --- NEW FILE: zCleanM --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 52 4 2 52 0 4 2 -4669.11 --- NEW FILE: zCleanN --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 52 4 2 52 0 4 2 -4681.37 --- NEW FILE: zMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 52 4 2 52 0 4 2 -4726.60 ICASE NB MFLOP ROUT AUTHOR 241 52 -9052.53 "ATL_dmm4x2x128_sse2.c" "Whaley & Voronenko" --- NEW FILE: zNB --- 1 52 --- NEW FILE: zNCNB --- 32 --- NEW FILE: zXover.h --- #ifndef ZXOVER_H #define ZXOVER_H #define ATL_3NB 156 #define NN_MNK_M 75088 #define NN_MNK_N 75088 #define NN_MNK_MN 27040 #define NN_MNK_K 5200 #define NN_MNK_GE 54872 #define NT_MNK_M 46800 #define NT_MNK_N 75088 #define NT_MNK_MN 27040 #define NT_MNK_K 5200 #define NT_MNK_GE 27000 #define TN_MNK_M 75088 #define TN_MNK_N 75088 #define TN_MNK_MN 27040 #define TN_MNK_K 5200 #define TN_MNK_GE 54872 #define TT_MNK_M 75088 #define TT_MNK_N 75088 #define TT_MNK_MN 27040 #define TT_MNK_K 5200 #define TT_MNK_GE 27000 #define C2R_K 1000 #endif --- NEW FILE: zbestNN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 32 4 2 1 0 4 2 -4110.88 --- NEW FILE: zbestNN_0x0x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 32 4 2 32 0 4 2 -4547.11 --- NEW FILE: zbestNN_32x32x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 513 32 4 2 32 0 4 2 -4557.44 --- NEW FILE: zbestNT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 32 4 2 1 0 4 2 -3905.09 --- NEW FILE: zbestNT_0x0x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 8 513 32 4 2 32 0 4 2 -4209.77 --- NEW FILE: zbestNT_32x32x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 32 4 2 32 0 4 2 -4217.71 --- NEW FILE: zbestTN_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 32 4 2 1 0 4 2 -4100.09 --- NEW FILE: zbestTN_0x0x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 2 513 32 4 2 32 0 4 2 -4346.03 --- NEW FILE: zbestTN_32x32x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 32 4 2 32 0 4 2 -4473.54 --- NEW FILE: zbestTT_0x0x0 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 513 32 4 2 1 0 4 2 -4054.40 --- NEW FILE: zbestTT_0x0x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 32 4 2 32 0 4 2 -4297.57 --- NEW FILE: zbestTT_32x32x32 --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 4 513 32 4 2 32 0 4 2 -4439.63 --- NEW FILE: zgMMRES --- MULADD LAT PREF NB MU NU KU FFTCH IFTCH NFTCH MFLOP 1 5 513 52 4 2 52 1 5 1 -4428.19 --- NEW FILE: zuCleanK --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 327 1 52 46 26 16 -3700.84 ATL_dmm4x1x90_x87.c 2 241 1 52 46 26 16 -8242.57 ATL_dmm4x2x128_sse2.c --- NEW FILE: zuCleanKF --- MULT ICASE FIXED NB 1 2 241 1 52 --- NEW FILE: zuCleanM --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 2 243 0 52 46 26 16 -8189.63 ATL_dmm2x2x128_sse2.c 4 241 0 52 44 24 16 -8488.88 ATL_dmm4x2x128_sse2.c --- NEW FILE: zuCleanMF --- MULT ICASE FIXED NB 2 2 243 0 52 4 241 0 52 --- NEW FILE: zuCleanN --- MULT ICASE FIXED NB NB0 NB1 NB2 MFLOP ROUT 2 1 327 0 52 46 26 16 -3884.64 ATL_dmm4x1x90_x87.c 2 241 0 52 46 26 16 -8672.34 ATL_dmm4x2x128_sse2.c --- NEW FILE: zuCleanNF --- MULT ICASE FIXED NB 1 2 241 0 52 --- NEW FILE: zuMMRES --- CASE NB MFLOP ROUTINE 241 52 -9052.53 "ATL_dmm4x2x128_sse2.c" "Whaley & Voronenko" |