The following benchmarks were designed and
performed by Roman "GoodOk" Gudchenko. The evaluated systems are:
- Intel Celeron 633 (SDRAM)
- Intel Pentium 4 2000 (DDR)
- Intel Pentium 4 1500 (SDRAM)
- AMD Athlon 700 (DDR)
Intel Celeron 633 (SDRAM)
Operation = Sum (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 28,03 us 47,90 us 47,71 us
4 55,76 us 47,53 us 47,98 us
8 132,52 us 72,88 us 55,70 us
16 243,39 us 122,92 us 68,38 us
32 450,35 us 243,43 us 99,86 us
64 862,69 us 445,65 us 164,59 us
128 1,80 ms 850,08 us 316,09 us
256 3,43 ms 1,66 ms 565,70 us
512 6,82 ms 3,28 ms 1,07 ms
1024 13,58 ms 6,52 ms 2,11 ms
2048 27,13 ms 13,12 ms 4,31 ms
4096 54,93 ms 26,01 ms 15,37 ms
8192 111,92 ms 52,08 ms 27,59 ms
----------------------------------------------------
Operation = Sub (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 27,25 us 44,23 us 45,29 us
4 55,62 us 43,94 us 45,06 us
8 143,91 us 65,43 us 52,91 us
16 249,63 us 109,30 us 67,09 us
32 442,38 us 217,03 us 98,79 us
64 837,52 us 395,13 us 161,98 us
128 1,48 ms 749,41 us 313,54 us
256 2,92 ms 1,46 ms 566,29 us
512 5,86 ms 2,91 ms 1,42 ms
1024 12,43 ms 5,83 ms 2,08 ms
2048 24,61 ms 11,48 ms 4,45 ms
4096 54,78 ms 27,37 ms 13,21 ms
8192 111,99 ms 55,84 ms 28,58 ms
----------------------------------------------------
Operation = Div (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 31,86 us 54,73 us 64,29 us
4 60,16 us 78,09 us 63,91 us
8 153,36 us 128,02 us 89,08 us
16 262,55 us 234,47 us 143,44 us
32 492,99 us 463,61 us 242,13 us
64 1,02 ms 872,62 us 450,69 us
128 1,86 ms 1,69 ms 869,19 us
256 3,67 ms 3,30 ms 1,70 ms
512 7,49 ms 6,54 ms 3,32 ms
1024 14,75 ms 13,03 ms 6,53 ms
2048 29,49 ms 26,08 ms 13,14 ms
4096 61,98 ms 53,05 ms 26,06 ms
8192 124,32 ms 105,96 ms 52,92 ms
----------------------------------------------------
Operation = Max (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 30,78 us 57,42 us 49,84 us
4 55,67 us 57,33 us 49,81 us
8 123,49 us 81,10 us 58,02 us
16 216,94 us 133,49 us 99,04 us
32 408,34 us 250,70 us 102,10 us
64 767,76 us 453,91 us 165,20 us
128 1,48 ms 858,71 us 355,77 us
256 2,93 ms 1,72 ms 569,71 us
512 5,75 ms 3,29 ms 1,08 ms
1024 11,42 ms 6,53 ms 2,11 ms
2048 24,02 ms 13,26 ms 4,44 ms
4096 50,60 ms 29,34 ms 15,37 ms
8192 102,69 ms 58,60 ms 26,67 ms
----------------------------------------------------
Operation = Inner Sum
----------------------------------------------------
Length Classic x87 SIMD
2 26,16 us 44,09 us 52,26 us
4 46,16 us 44,05 us 52,03 us
8 129,66 us 65,07 us 57,40 us
16 230,78 us 116,12 us 83,63 us
32 432,98 us 251,84 us 127,22 us
64 837,40 us 466,69 us 158,32 us
128 1,65 ms 896,40 us 249,43 us
256 3,26 ms 1,76 ms 426,91 us
512 6,51 ms 3,50 ms 790,36 us
1024 13,00 ms 7,01 ms 1,51 ms
2048 25,98 ms 13,83 ms 2,97 ms
4096 51,99 ms 27,71 ms 5,95 ms
8192 105,51 ms 55,29 ms 12,50 ms
----------------------------------------------------
Operation = NormL2
----------------------------------------------------
Length Classic x87 SIMD
2 39,70 us 91,16 us 63,88 us
4 72,90 us 90,96 us 63,61 us
8 167,55 us 162,05 us 74,42 us
16 326,40 us 263,21 us 99,64 us
32 577,44 us 465,20 us 171,52 us
64 1,08 ms 887,26 us 263,62 us
128 2,15 ms 1,68 ms 469,76 us
256 4,28 ms 3,32 ms 830,00 us
512 8,52 ms 6,56 ms 1,59 ms
1024 16,81 ms 13,08 ms 3,12 ms
2048 33,69 ms 26,19 ms 6,25 ms
4096 69,94 ms 52,38 ms 13,63 ms
8192 139,77 ms 104,88 ms 27,28 ms
----------------------------------------------------
Operation = Lerp (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 48,84 us 70,81 us 42,91 us
4 94,98 us 70,86 us 42,92 us
8 213,79 us 125,64 us 56,75 us
16 389,23 us 183,61 us 83,92 us
32 760,33 us 353,49 us 155,46 us
64 1,47 ms 656,86 us 262,88 us
128 2,91 ms 1,27 ms 476,17 us
256 5,78 ms 2,50 ms 900,36 us
512 11,59 ms 4,93 ms 1,75 ms
1024 23,31 ms 9,78 ms 3,46 ms
2048 46,95 ms 19,57 ms 7,06 ms
4096 94,98 ms 40,99 ms 20,29 ms
8192 191,48 ms 84,17 ms 37,89 ms
----------------------------------------------------
Operation = Sum (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 37,74 us 51,95 us 54,76 us
4 70,78 us 50,72 us 54,58 us
8 137,54 us 76,06 us 65,77 us
16 276,55 us 126,65 us 79,89 us
32 545,13 us 249,78 us 113,24 us
64 1,13 ms 451,99 us 171,69 us
128 2,17 ms 856,44 us 320,12 us
256 4,33 ms 1,67 ms 572,71 us
512 8,72 ms 3,28 ms 1,08 ms
1024 17,42 ms 6,54 ms 2,09 ms
2048 36,23 ms 13,54 ms 8,19 ms
4096 71,99 ms 30,20 ms 18,33 ms
8192 149,05 ms 151,76 ms 41,47 ms
----------------------------------------------------
Operation = Sub (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 37,95 us 50,29 us 53,06 us
4 71,53 us 49,69 us 52,82 us
8 174,00 us 73,46 us 59,06 us
16 316,11 us 117,80 us 79,72 us
32 589,30 us 224,94 us 108,10 us
64 1,16 ms 401,88 us 174,53 us
128 2,19 ms 756,07 us 326,11 us
256 4,44 ms 1,46 ms 591,44 us
512 8,91 ms 2,88 ms 1,18 ms
1024 17,71 ms 5,92 ms 2,19 ms
2048 35,95 ms 13,06 ms 8,17 ms
4096 71,43 ms 31,34 ms 22,12 ms
8192 145,10 ms 66,21 ms 45,06 ms
----------------------------------------------------
Operation = Mult (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 38,54 us 51,15 us 53,14 us
4 74,13 us 50,25 us 52,27 us
8 163,23 us 71,32 us 61,83 us
16 314,87 us 115,62 us 74,40 us
32 599,20 us 222,17 us 110,75 us
64 1,18 ms 399,49 us 175,56 us
128 2,30 ms 752,54 us 327,33 us
256 4,55 ms 1,47 ms 586,42 us
512 9,03 ms 2,88 ms 1,20 ms
1024 18,09 ms 7,31 ms 2,14 ms
2048 37,90 ms 13,98 ms 8,24 ms
4096 74,74 ms 31,30 ms 23,13 ms
8192 150,11 ms 120,74 ms 40,10 ms
----------------------------------------------------
Operation = Div (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 117,21 us 154,45 us 81,70 us
4 234,32 us 276,30 us 95,14 us
8 549,91 us 508,53 us 156,60 us
16 1,09 ms 978,73 us 262,90 us
32 2,14 ms 1,93 ms 476,39 us
64 4,31 ms 3,88 ms 905,60 us
128 8,54 ms 7,55 ms 1,78 ms
256 17,09 ms 15,04 ms 3,50 ms
512 34,14 ms 30,02 ms 6,94 ms
1024 68,12 ms 60,03 ms 13,84 ms
2048 138,86 ms 120,02 ms 28,50 ms
4096 273,71 ms 239,92 ms 55,41 ms
8192 550,75 ms 482,10 ms 111,01 ms
----------------------------------------------------
Operation = Max (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 53,70 us 70,75 us 53,20 us
4 100,42 us 71,65 us 53,28 us
8 220,53 us 104,28 us 59,49 us
16 485,89 us 178,37 us 80,23 us
32 1,01 ms 400,59 us 114,68 us
64 2,11 ms 996,34 us 173,28 us
128 4,85 ms 2,15 ms 319,95 us
256 9,94 ms 4,63 ms 572,92 us
512 20,57 ms 10,14 ms 1,16 ms
1024 39,67 ms 20,69 ms 2,09 ms
2048 79,22 ms 43,81 ms 8,21 ms
4096 157,23 ms 88,69 ms 22,17 ms
8192 316,12 ms 174,79 ms 44,78 ms
----------------------------------------------------
Intel Pentium 4 2000 (DDR)
Operation = Sum (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 20,17 us 16,19 us 11,16 us
4 15,24 us 15,16 us 10,15 us
8 33,25 us 21,22 us 14,16 us
16 65,40 us 36,85 us 21,21 us
32 110,57 us 67,43 us 31,24 us
64 206,94 us 140,17 us 54,41 us
128 440,42 us 303,87 us 111,57 us
256 871,63 us 630,42 us 200,98 us
512 1,72 ms 1,28 ms 398,82 us
1024 5,72 ms 2,42 ms 774,78 us
2048 7,47 ms 4,91 ms 1,97 ms
4096 15,36 ms 10,16 ms 4,35 ms
8192 28,97 ms 18,31 ms 9,97 ms
----------------------------------------------------
Operation = Sub (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 11,46 us 16,31 us 11,74 us
4 17,23 us 15,15 us 10,15 us
8 34,26 us 21,69 us 14,17 us
16 65,38 us 36,81 us 21,22 us
32 110,56 us 68,93 us 34,23 us
64 206,94 us 130,23 us 56,35 us
128 440,39 us 303,85 us 104,54 us
256 875,09 us 630,44 us 206,90 us
512 1,78 ms 1,28 ms 398,89 us
1024 3,61 ms 2,54 ms 778,13 us
2048 7,42 ms 5,10 ms 1,97 ms
4096 15,16 ms 10,23 ms 4,30 ms
8192 30,64 ms 18,46 ms 9,99 ms
----------------------------------------------------
Operation = Div (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 23,17 us 23,45 us 28,24 us
4 46,26 us 46,33 us 28,08 us
8 92,47 us 92,59 us 39,25 us
16 184,80 us 185,01 us 78,42 us
32 369,53 us 369,73 us 156,82 us
64 739,01 us 739,29 us 328,87 us
128 1,48 ms 1,54 ms 626,68 us
256 2,96 ms 3,01 ms 1,25 ms
512 5,91 ms 5,98 ms 2,51 ms
1024 12,90 ms 11,91 ms 5,01 ms
2048 23,69 ms 23,72 ms 10,03 ms
4096 56,97 ms 47,44 ms 20,06 ms
8192 110,15 ms 94,70 ms 40,17 ms
----------------------------------------------------
Operation = Max (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 13,14 us 25,73 us 13,80 us
4 19,37 us 21,26 us 10,13 us
8 35,49 us 31,56 us 12,15 us
16 92,81 us 51,82 us 21,23 us
32 151,14 us 112,11 us 32,25 us
64 279,49 us 215,20 us 63,36 us
128 536,69 us 409,91 us 104,54 us
256 1,05 ms 669,54 us 195,95 us
512 2,08 ms 1,32 ms 398,76 us
1024 4,16 ms 2,66 ms 775,02 us
2048 8,46 ms 5,19 ms 1,97 ms
4096 16,98 ms 11,41 ms 4,31 ms
8192 34,08 ms 19,60 ms 9,96 ms
----------------------------------------------------
Operation = Inner Sum
----------------------------------------------------
Length Classic x87 SIMD
2 10,66 us 15,47 us 17,98 us
4 36,96 us 15,17 us 18,65 us
8 67,85 us 25,80 us 19,42 us
16 118,71 us 58,14 us 23,31 us
32 351,66 us 111,70 us 31,27 us
64 806,41 us 328,93 us 49,29 us
128 1,74 ms 686,30 us 79,44 us
256 3,62 ms 1,40 ms 143,66 us
512 7,12 ms 2,83 ms 272,39 us
1024 14,79 ms 5,38 ms 529,19 us
2048 29,65 ms 10,94 ms 1,09 ms
4096 58,46 ms 21,17 ms 2,42 ms
8192 115,27 ms 43,67 ms 4,14 ms
----------------------------------------------------
Operation = NormL2
----------------------------------------------------
Length Classic x87 SIMD
2 14,66 us 42,30 us 21,35 us
4 31,24 us 41,13 us 22,06 us
8 72,11 us 73,70 us 24,82 us
16 149,29 us 188,13 us 30,19 us
32 387,17 us 419,04 us 40,81 us
64 853,02 us 880,83 us 66,49 us
128 1,81 ms 1,80 ms 106,63 us
256 3,60 ms 3,55 ms 186,96 us
512 7,35 ms 7,35 ms 350,73 us
1024 14,88 ms 14,25 ms 775,09 us
2048 29,84 ms 28,50 ms 1,52 ms
4096 59,81 ms 57,62 ms 3,03 ms
8192 119,43 ms 112,66 ms 5,98 ms
----------------------------------------------------
Operation = Lerp (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 17,17 us 22,73 us 22,46 us
4 26,73 us 22,24 us 22,15 us
8 61,60 us 33,16 us 32,65 us
16 105,06 us 52,41 us 52,25 us
32 219,04 us 94,08 us 93,95 us
64 426,85 us 173,97 us 173,79 us
128 868,89 us 396,24 us 402,78 us
256 1,75 ms 835,35 us 774,17 us
512 3,44 ms 1,64 ms 1,62 ms
1024 6,92 ms 3,26 ms 3,22 ms
2048 13,82 ms 6,23 ms 6,35 ms
4096 27,96 ms 12,13 ms 14,81 ms
8192 55,86 ms 31,31 ms 35,14 ms
----------------------------------------------------
Operation = Sum (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 12,63 us 18,62 us 11,82 us
4 24,72 us 17,18 us 11,16 us
8 48,80 us 23,23 us 13,18 us
16 126,01 us 36,86 us 21,21 us
32 198,91 us 73,00 us 35,21 us
64 391,69 us 128,96 us 57,35 us
128 777,24 us 232,16 us 105,55 us
256 1,56 ms 440,98 us 201,96 us
512 3,09 ms 849,45 us 374,81 us
1024 6,31 ms 2,02 ms 989,65 us
2048 13,45 ms 3,44 ms 1,94 ms
4096 27,16 ms 8,77 ms 3,80 ms
8192 54,52 ms 17,57 ms 13,02 ms
----------------------------------------------------
Operation = Sub (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 12,65 us 18,54 us 12,55 us
4 24,67 us 16,18 us 11,15 us
8 54,29 us 22,21 us 13,14 us
16 96,97 us 36,82 us 21,21 us
32 198,86 us 65,41 us 31,26 us
64 391,63 us 130,65 us 53,39 us
128 783,63 us 231,66 us 105,56 us
256 1,56 ms 423,53 us 201,97 us
512 3,11 ms 868,31 us 375,17 us
1024 7,23 ms 2,03 ms 999,31 us
2048 12,58 ms 3,80 ms 1,94 ms
4096 27,16 ms 8,78 ms 3,86 ms
8192 54,18 ms 17,55 ms 13,03 ms
----------------------------------------------------
Operation = Mult (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 21,16 us 18,63 us 11,48 us
4 24,73 us 18,19 us 13,13 us
8 48,79 us 24,22 us 15,17 us
16 97,00 us 38,34 us 23,25 us
32 198,91 us 65,56 us 33,25 us
64 391,71 us 134,14 us 57,41 us
128 777,23 us 246,11 us 105,55 us
256 1,55 ms 506,05 us 196,95 us
512 3,13 ms 838,75 us 374,81 us
1024 6,48 ms 2,15 ms 986,66 us
2048 12,47 ms 4,25 ms 1,94 ms
4096 27,21 ms 9,05 ms 3,95 ms
8192 54,55 ms 17,82 ms 12,49 ms
----------------------------------------------------
Operation = Div (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 43,26 us 43,56 us 20,13 us
4 86,42 us 86,48 us 19,69 us
8 172,80 us 172,94 us 39,24 us
16 345,43 us 345,56 us 78,50 us
32 690,84 us 690,99 us 156,85 us
64 1,45 ms 1,38 ms 313,36 us
128 2,76 ms 2,82 ms 626,61 us
256 5,53 ms 5,59 ms 1,25 ms
512 11,06 ms 11,11 ms 2,52 ms
1024 22,12 ms 22,17 ms 5,02 ms
2048 44,26 ms 44,29 ms 10,02 ms
4096 88,48 ms 88,53 ms 20,05 ms
8192 177,29 ms 177,07 ms 40,20 ms
----------------------------------------------------
Operation = Max (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 14,85 us 27,73 us 11,71 us
4 27,29 us 24,52 us 11,14 us
8 73,65 us 37,43 us 13,20 us
16 108,98 us 66,81 us 21,22 us
32 314,83 us 104,74 us 33,28 us
64 670,43 us 309,56 us 64,36 us
128 1,27 ms 483,35 us 105,88 us
256 3,71 ms 1,14 ms 202,01 us
512 8,38 ms 2,41 ms 373,84 us
1024 18,38 ms 5,57 ms 992,62 us
2048 36,10 ms 14,74 ms 1,94 ms
4096 70,43 ms 48,85 ms 3,89 ms
8192 141,47 ms 107,67 ms 12,90 ms
----------------------------------------------------
Intel Pentium 4 1500
(SDRAM)
Operation = Sum (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 15,12 us 21,50 us 14,75 us
4 22,93 us 21,35 us 14,75 us
8 43,79 us 30,62 us 17,41 us
16 86,14 us 51,29 us 26,59 us
32 145,75 us 86,21 us 42,44 us
64 272,63 us 164,10 us 74,22 us
128 562,52 us 412,31 us 140,38 us
256 1,13 ms 756,83 us 264,78 us
512 2,55 ms 1,58 ms 491,02 us
1024 4,60 ms 3,31 ms 974,14 us
2048 10,08 ms 6,66 ms 2,36 ms
4096 20,76 ms 13,76 ms 4,91 ms
8192 38,39 ms 24,77 ms 9,82 ms
65536 887,09 ms 890,08 ms 906,23 ms
----------------------------------------------------
Operation = Sub (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 25,89 us 21,77 us 15,01 us
4 21,32 us 19,98 us 14,58 us
8 37,27 us 30,60 us 19,05 us
16 86,15 us 51,29 us 28,00 us
32 146,44 us 89,97 us 43,90 us
64 272,81 us 164,94 us 75,87 us
128 584,84 us 397,64 us 137,96 us
256 1,14 ms 837,64 us 265,00 us
512 2,30 ms 1,51 ms 491,33 us
1024 4,60 ms 2,89 ms 974,55 us
2048 9,37 ms 6,70 ms 2,44 ms
4096 19,89 ms 13,27 ms 4,81 ms
8192 40,99 ms 25,82 ms 10,11 ms
65536 888,23 ms 893,44 ms 888,34 ms
----------------------------------------------------
Operation = Div (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 30,52 us 30,95 us 36,83 us
4 60,99 us 61,07 us 36,84 us
8 121,75 us 122,00 us 52,07 us
16 243,46 us 243,62 us 103,44 us
32 486,84 us 487,10 us 267,35 us
64 973,61 us 973,95 us 439,26 us
128 1,95 ms 2,02 ms 825,75 us
256 3,89 ms 3,97 ms 1,68 ms
512 7,79 ms 7,89 ms 3,30 ms
1024 15,64 ms 15,66 ms 6,64 ms
2048 31,18 ms 31,25 ms 13,25 ms
4096 71,38 ms 62,61 ms 26,43 ms
8192 124,75 ms 127,76 ms 52,90 ms
65536 1,13 sec 1,08 sec 889,60 ms
----------------------------------------------------
Operation = Max (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 16,79 us 28,37 us 14,97 us
4 27,10 us 28,50 us 14,72 us
8 54,99 us 40,45 us 18,71 us
16 114,24 us 76,23 us 26,59 us
32 199,25 us 137,47 us 41,12 us
64 452,01 us 284,50 us 72,89 us
128 709,11 us 460,10 us 136,40 us
256 1,39 ms 1,13 ms 263,39 us
512 2,75 ms 2,10 ms 489,63 us
1024 6,52 ms 3,49 ms 972,32 us
2048 11,15 ms 6,46 ms 2,40 ms
4096 23,65 ms 13,86 ms 4,82 ms
8192 47,02 ms 30,18 ms 10,08 ms
65536 542,67 ms 545,52 ms 892,87 ms
----------------------------------------------------
Operation = Inner Sum
----------------------------------------------------
Length Classic x87 SIMD
2 13,86 us 20,40 us 25,52 us
4 46,72 us 19,40 us 24,53 us
8 74,69 us 32,43 us 25,24 us
16 164,24 us 68,76 us 43,75 us
32 463,12 us 153,54 us 53,10 us
64 1,06 ms 354,99 us 64,11 us
128 2,31 ms 861,98 us 116,59 us
256 4,75 ms 1,73 ms 201,30 us
512 9,68 ms 3,47 ms 359,13 us
1024 19,53 ms 6,94 ms 697,66 us
2048 39,17 ms 14,36 ms 1,71 ms
4096 75,60 ms 28,87 ms 3,34 ms
8192 150,85 ms 57,32 ms 6,52 ms
65536 1,21 sec 461,73 ms 111,47 ms
----------------------------------------------------
Operation = NormL2
----------------------------------------------------
Length Classic x87 SIMD
2 19,30 us 56,13 us 28,28 us
4 40,72 us 54,80 us 28,23 us
8 85,17 us 97,42 us 33,14 us
16 212,13 us 229,15 us 39,89 us
32 510,30 us 577,19 us 53,64 us
64 1,12 ms 1,16 ms 88,00 us
128 2,36 ms 2,38 ms 140,91 us
256 4,83 ms 4,81 ms 246,75 us
512 9,87 ms 9,68 ms 458,45 us
1024 19,58 ms 19,06 ms 977,03 us
2048 39,23 ms 38,72 ms 2,02 ms
4096 78,65 ms 76,46 ms 3,97 ms
8192 157,16 ms 149,22 ms 7,92 ms
65536 1,28 sec 1,23 sec 556,30 ms
----------------------------------------------------
Operation = Lerp (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 29,21 us 30,12 us 22,36 us
4 35,25 us 27,42 us 21,00 us
8 69,66 us 42,86 us 22,25 us
16 138,44 us 68,05 us 31,03 us
32 286,51 us 122,93 us 52,12 us
64 561,72 us 246,52 us 105,09 us
128 1,14 ms 541,60 us 151,34 us
256 2,30 ms 1,06 ms 278,35 us
512 4,56 ms 2,02 ms 532,36 us
1024 9,12 ms 4,18 ms 1,07 ms
2048 18,48 ms 8,47 ms 2,40 ms
4096 39,77 ms 19,72 ms 5,67 ms
8192 76,65 ms 41,35 ms 9,96 ms
65536 900,44 ms 888,15 ms 896,16 ms
----------------------------------------------------
Operation = Sum (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 27,21 us 24,54 us 14,88 us
4 32,50 us 21,59 us 17,47 us
8 64,28 us 29,53 us 20,18 us
16 134,99 us 50,65 us 28,09 us
32 262,03 us 86,32 us 41,28 us
64 515,98 us 160,43 us 70,44 us
128 1,03 ms 305,13 us 137,92 us
256 2,04 ms 550,08 us 264,85 us
512 4,09 ms 1,07 ms 491,17 us
1024 8,24 ms 3,14 ms 1,19 ms
2048 16,61 ms 4,81 ms 2,93 ms
4096 35,89 ms 11,60 ms 5,05 ms
8192 65,62 ms 39,83 ms 10,16 ms
65536 1,24 sec 1,17 sec 1,19 sec
----------------------------------------------------
Operation = Sub (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 27,22 us 24,43 us 15,26 us
4 40,43 us 39,98 us 17,79 us
8 71,52 us 36,07 us 20,32 us
16 127,74 us 54,62 us 30,95 us
32 261,99 us 88,98 us 44,25 us
64 515,96 us 167,67 us 75,88 us
128 1,03 ms 308,01 us 142,16 us
256 2,06 ms 567,36 us 269,15 us
512 4,07 ms 1,08 ms 492,61 us
1024 8,25 ms 3,01 ms 1,15 ms
2048 16,48 ms 6,26 ms 2,46 ms
4096 36,69 ms 12,37 ms 5,16 ms
8192 71,60 ms 23,16 ms 9,27 ms
65536 1,18 sec 1,20 sec 1,21 sec
----------------------------------------------------
Operation = Mult (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 27,21 us 24,49 us 1,21 ms
4 32,58 us 21,44 us 1,16 ms
8 64,47 us 29,46 us 2,38 ms
16 128,14 us 49,43 us 31,06 us
32 262,37 us 86,25 us 44,21 us
64 516,63 us 160,36 us 75,91 us
128 1,03 ms 283,62 us 139,43 us
256 2,04 ms 649,61 us 266,42 us
512 4,07 ms 1,13 ms 492,75 us
1024 8,24 ms 3,13 ms 1,19 ms
2048 16,56 ms 4,71 ms 2,88 ms
4096 35,98 ms 11,95 ms 5,03 ms
8192 65,63 ms 20,76 ms 9,87 ms
65536 1,24 sec 1,17 sec 1,19 sec
----------------------------------------------------
Operation = Div (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 56,98 us 57,53 us 26,47 us
4 114,03 us 114,24 us 26,33 us
8 227,73 us 227,82 us 52,06 us
16 455,33 us 455,33 us 103,66 us
32 910,39 us 910,35 us 206,81 us
64 1,84 ms 1,82 ms 413,16 us
128 3,64 ms 3,72 ms 825,88 us
256 7,30 ms 7,36 ms 1,65 ms
512 14,56 ms 14,69 ms 3,31 ms
1024 29,15 ms 29,24 ms 6,60 ms
2048 58,33 ms 58,35 ms 13,23 ms
4096 116,87 ms 116,65 ms 26,43 ms
8192 233,19 ms 233,26 ms 52,91 ms
65536 2,00 sec 1,94 sec 1,21 sec
----------------------------------------------------
Operation = Max (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 45,14 us 39,13 us 15,13 us
4 36,07 us 32,29 us 14,85 us
8 83,45 us 49,59 us 17,53 us
16 164,87 us 128,12 us 28,08 us
32 387,51 us 182,44 us 42,58 us
64 823,94 us 337,93 us 74,36 us
128 1,77 ms 736,11 us 139,17 us
256 4,03 ms 1,48 ms 264,88 us
512 10,85 ms 2,89 ms 511,24 us
1024 23,63 ms 6,62 ms 1,19 ms
2048 46,75 ms 20,18 ms 2,93 ms
4096 90,67 ms 63,77 ms 5,05 ms
8192 177,87 ms 211,66 ms 10,22 ms
65536 2,34 sec 2,05 sec 1,18 sec
----------------------------------------------------
AMD Athlon 700
(DDR)
Operation = Sum (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 37,44 us 38,61 us 26,02 us
4 74,01 us 38,42 us 25,66 us
8 147,68 us 54,25 us 39,99 us
16 345,74 us 74,01 us 44,07 us
32 661,44 us 125,03 us 66,87 us
64 1,23 ms 260,85 us 112,34 us
128 2,42 ms 464,79 us 203,06 us
256 4,79 ms 881,78 us 411,02 us
512 9,63 ms 1,86 ms 1,14 ms
1024 19,25 ms 3,38 ms 1,61 ms
2048 38,46 ms 6,64 ms 3,17 ms
4096 77,11 ms 17,62 ms 6,60 ms
8192 153,52 ms 26,90 ms 14,92 ms
----------------------------------------------------
Operation = Sub (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 37,36 us 38,43 us 25,39 us
4 74,04 us 38,45 us 24,96 us
8 147,91 us 50,07 us 32,89 us
16 328,91 us 72,65 us 45,53 us
32 626,36 us 117,90 us 68,40 us
64 1,22 ms 241,05 us 113,75 us
128 2,41 ms 422,31 us 204,34 us
256 4,90 ms 784,92 us 411,05 us
512 9,57 ms 1,53 ms 1,34 ms
1024 19,23 ms 3,04 ms 2,21 ms
2048 38,43 ms 5,94 ms 3,67 ms
4096 77,17 ms 14,85 ms 6,70 ms
8192 153,62 ms 24,02 ms 13,25 ms
----------------------------------------------------
Operation = Div (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 40,05 us 159,11 us 42,48 us
4 77,08 us 187,08 us 42,01 us
8 150,79 us 137,28 us 65,81 us
16 337,49 us 227,23 us 119,77 us
32 634,80 us 408,59 us 232,85 us
64 1,23 ms 809,28 us 453,83 us
128 2,42 ms 1,52 ms 915,54 us
256 4,82 ms 2,97 ms 1,86 ms
512 9,57 ms 5,89 ms 3,66 ms
1024 19,24 ms 11,86 ms 7,27 ms
2048 38,44 ms 23,48 ms 14,51 ms
4096 77,68 ms 46,83 ms 29,10 ms
8192 154,97 ms 93,75 ms 57,76 ms
----------------------------------------------------
Operation = Max (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 46,42 us 67,31 us 26,59 us
4 85,44 us 67,00 us 25,77 us
8 184,91 us 108,22 us 32,99 us
16 335,17 us 208,75 us 44,20 us
32 632,90 us 367,80 us 66,96 us
64 1,25 ms 735,17 us 112,35 us
128 2,49 ms 1,32 ms 203,06 us
256 4,80 ms 2,63 ms 408,37 us
512 9,58 ms 5,15 ms 955,08 us
1024 19,22 ms 10,27 ms 1,80 ms
2048 38,45 ms 20,51 ms 3,68 ms
4096 77,46 ms 41,16 ms 6,63 ms
8192 153,92 ms 82,22 ms 14,93 ms
----------------------------------------------------
Operation = Inner Sum
----------------------------------------------------
Length Classic x87 SIMD
2 21,44 us 31,80 us 32,55 us
4 45,36 us 31,36 us 33,01 us
8 96,89 us 54,94 us 40,33 us
16 307,58 us 98,32 us 57,65 us
32 575,20 us 221,67 us 102,34 us
64 1,08 ms 591,01 us 193,73 us
128 2,10 ms 1,20 ms 377,19 us
256 4,13 ms 2,33 ms 772,31 us
512 8,22 ms 4,63 ms 1,50 ms
1024 16,42 ms 9,20 ms 2,95 ms
2048 32,67 ms 18,38 ms 5,85 ms
4096 65,24 ms 36,69 ms 11,66 ms
8192 130,36 ms 73,37 ms 23,29 ms
----------------------------------------------------
Operation = NormL2
----------------------------------------------------
Length Classic x87 SIMD
2 38,55 us 79,22 us 35,75 us
4 75,59 us 77,96 us 35,27 us
8 152,35 us 136,46 us 49,20 us
16 363,07 us 345,86 us 77,79 us
32 646,28 us 617,88 us 107,95 us
64 1,25 ms 1,17 ms 240,98 us
128 2,61 ms 2,29 ms 419,45 us
256 4,87 ms 4,61 ms 783,47 us
512 9,67 ms 9,29 ms 1,51 ms
1024 20,55 ms 17,92 ms 2,97 ms
2048 38,55 ms 36,21 ms 5,86 ms
4096 77,10 ms 71,98 ms 11,72 ms
8192 164,32 ms 143,98 ms 23,53 ms
----------------------------------------------------
Operation = Lerp (single operand)
----------------------------------------------------
Length Classic x87 SIMD
2 57,00 us 47,62 us 28,96 us
4 113,74 us 48,32 us 28,69 us
8 232,89 us 66,94 us 38,57 us
16 490,43 us 104,77 us 58,55 us
32 949,34 us 183,14 us 98,08 us
64 1,87 ms 379,77 us 211,24 us
128 3,71 ms 695,62 us 369,88 us
256 7,39 ms 1,31 ms 688,58 us
512 14,81 ms 2,55 ms 1,50 ms
1024 29,52 ms 5,04 ms 2,95 ms
2048 58,99 ms 10,42 ms 5,14 ms
4096 118,21 ms 21,87 ms 10,41 ms
8192 235,87 ms 41,37 ms 20,92 ms
----------------------------------------------------
Operation = Sum (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 52,60 us 41,85 us 34,76 us
4 106,52 us 42,68 us 31,39 us
8 208,66 us 57,12 us 42,86 us
16 437,92 us 82,59 us 52,85 us
32 851,44 us 135,02 us 78,29 us
64 1,69 ms 268,02 us 132,04 us
128 3,33 ms 471,94 us 221,30 us
256 6,65 ms 881,30 us 440,79 us
512 13,29 ms 1,87 ms 997,39 us
1024 26,56 ms 4,06 ms 2,19 ms
2048 53,08 ms 6,97 ms 3,14 ms
4096 106,97 ms 16,30 ms 7,36 ms
8192 223,58 ms 62,86 ms 31,90 ms
----------------------------------------------------
Operation = Sub (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 52,79 us 41,64 us 33,25 us
4 106,53 us 41,27 us 34,22 us
8 208,52 us 54,23 us 41,41 us
16 437,82 us 79,80 us 52,88 us
32 851,41 us 125,23 us 82,35 us
64 1,68 ms 246,85 us 125,02 us
128 3,33 ms 439,91 us 224,06 us
256 6,64 ms 790,61 us 453,52 us
512 13,27 ms 1,88 ms 960,44 us
1024 26,56 ms 4,15 ms 2,14 ms
2048 53,09 ms 6,61 ms 3,19 ms
4096 107,15 ms 16,28 ms 7,49 ms
8192 223,16 ms 67,54 ms 32,28 ms
----------------------------------------------------
Operation = Mult (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 52,99 us 42,35 us 30,73 us
4 106,43 us 41,96 us 29,97 us
8 208,44 us 54,19 us 40,04 us
16 437,84 us 78,38 us 51,23 us
32 851,40 us 126,40 us 73,95 us
64 1,69 ms 258,08 us 136,38 us
128 3,33 ms 450,60 us 228,61 us
256 6,65 ms 835,91 us 456,46 us
512 13,30 ms 1,80 ms 915,28 us
1024 26,57 ms 4,06 ms 2,15 ms
2048 53,13 ms 7,32 ms 3,31 ms
4096 106,57 ms 16,58 ms 7,51 ms
8192 223,69 ms 64,71 ms 31,64 ms
----------------------------------------------------
Operation = Div (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 60,27 us 61,06 us 47,44 us
4 119,47 us 119,47 us 46,38 us
8 238,39 us 238,36 us 69,97 us
16 524,22 us 476,57 us 122,62 us
32 998,73 us 952,41 us 237,21 us
64 1,95 ms 1,94 ms 473,67 us
128 3,85 ms 3,85 ms 916,19 us
256 7,67 ms 7,64 ms 1,89 ms
512 15,29 ms 15,34 ms 3,70 ms
1024 30,67 ms 30,52 ms 7,21 ms
2048 61,11 ms 61,05 ms 14,66 ms
4096 122,53 ms 122,05 ms 28,81 ms
8192 264,39 ms 244,29 ms 61,54 ms
----------------------------------------------------
Operation = Max (Dual)
----------------------------------------------------
Length Classic x87 SIMD
2 72,52 us 91,52 us 33,51 us
4 146,29 us 91,00 us 34,21 us
8 323,30 us 137,78 us 41,49 us
16 683,12 us 224,87 us 51,26 us
32 1,43 ms 453,21 us 76,84 us
64 3,08 ms 840,30 us 125,05 us
128 6,74 ms 1,89 ms 224,19 us
256 14,09 ms 4,07 ms 442,38 us
512 27,40 ms 9,64 ms 945,01 us
1024 56,91 ms 22,93 ms 2,15 ms
2048 111,35 ms 51,06 ms 3,14 ms
4096 223,09 ms 108,13 ms 7,49 ms
8192 458,57 ms 228,41 ms 32,52 ms
----------------------------------------------------

|