Skip to content

Test_BenchMark_Win_5950X.md

Latest commit

 

History

History
52 lines (50 loc) · 4.92 KB

Test_BenchMark_Win_5950X.md

File metadata and controls

52 lines (50 loc) · 4.92 KB

GFloat Test And BenchMark

  • Test time : Wed Nov 16 16:07:10 2022
Operation System C++ Compiler version CPU Base Frequency
Windows 64-bit Visual Studio :1932 AMD Ryzen 9 5950X 16-Core Processor 3.39 GHz or 0 GHz
  • Performance: float vs GFloat, Call 1000000 times
  • Error : the relative error between cmath (double) and GFloat Math
Function input data range avg relative error max relative error max abs error float vs GFloat float / GFloat
Mul+Add [ -10000.0, 10000.0] 0.000029 % 4.347826 % 32.000000 0.53 vs 6.81 (ms) 0.08
Add [ -10000.0, 10000.0] 0.000004 % 0.000024 % 0.003906 0.51 vs 6.10 (ms) 0.08
Sub [ -10000.0, 10000.0] 0.000004 % 0.000024 % 0.003906 0.91 vs 8.80 (ms) 0.10
Mul [ -10000.0, 10000.0] 0.000009 % 0.000024 % 16.000000 0.67 vs 2.82 (ms) 0.24
Div [ -10000.0, 10000.0] 0.000009 % 0.000024 % 0.031250 0.88 vs 3.17 (ms) 0.28
Ceil [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 6.42 vs 3.52 (ms) 1.82
Floor [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 6.28 vs 1.94 (ms) 3.24
Whole [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.67 vs 5.27 (ms) 0.13
WholeFrac [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.44 vs 10.12 (ms) 0.04
Fraction [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.45 vs 5.89 (ms) 0.08
-() [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.37 vs 1.93 (ms) 0.19
>() [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.50 vs 3.95 (ms) 0.13
<() [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.50 vs 8.83 (ms) 0.06
>=() [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.49 vs 8.65 (ms) 0.06
<=() [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.57 vs 5.09 (ms) 0.11
Abs [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.40 vs 4.64 (ms) 0.09
Normalize32 [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.41 vs 1.88 (ms) 0.22
Normalize64 [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.39 vs 1.96 (ms) 0.20
FromInt [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.51 vs 2.01 (ms) 0.25
Fromfloat [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 0.39 vs 0.62 (ms) 0.62
CeilToInt [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 6.42 vs 3.36 (ms) 1.91
FloorToInt [ -10000.0, 10000.0] 0.000000 % 0.000000 % 0.000000 6.25 vs 3.10 (ms) 2.02
Sin [ -10000.0, 10000.0] 0.000099 % 4.559921 % 0.000005 9.37 vs 8.99 (ms) 1.04
Cos [ -10000.0, 10000.0] 0.001805 % 0.122642 % 0.000028 9.23 vs 10.18 (ms) 0.91
Tan [ -10000.0, 10000.0] 0.001776 % 4.562861 % 83.726562 8.84 vs 20.49 (ms) 0.43
ASin [ -1.0, 1.0] 0.031373 % 100.000000 % 0.049958 8.09 vs 15.70 (ms) 0.52
ACos [ -1.0, 1.0] 0.117627 % 98.413861 % 0.049958 7.80 vs 26.50 (ms) 0.29
ATan [ -10000.0, 10000.0] 0.003567 % 0.036536 % 0.000166 8.79 vs 23.00 (ms) 0.38
ATan(10,x) [ -10000.0, 10000.0] 0.006190 % 0.021383 % 0.000166 11.22 vs 27.85 (ms) 0.40
ATan(x,10) [ -10000.0, 10000.0] 0.007318 % 0.230349 % 0.000166 11.46 vs 26.37 (ms) 0.43
Sqrt [ 0.0, 10000.0] 0.000034 % 0.000072 % 0.000061 1.14 vs 19.29 (ms) 0.06
InvSqrt [ 0.0, 10000.0] 0.000026 % 0.000062 % 0.000183 1.91 vs 14.25 (ms) 0.13
Exp [ -20.0, 20.0] 0.001689 % 0.050400 % 19056.000000 2.34 vs 13.64 (ms) 0.17
Pow2 [ -30.0, 30.0] 0.001820 % 0.050354 % 72448.000000 6.42 vs 9.00 (ms) 0.71
Log_e [ 0.0, 10000.0] 0.001206 % 86.225453 % 0.000031 3.12 vs 5.97 (ms) 0.52
Log_2 [ 0.0, 10000.0] 0.001206 % 86.225884 % 0.000043 24.27 vs 5.67 (ms) 4.28
Log_10 [ 0.0, 10000.0] 0.001206 % 86.224927 % 0.000013 3.33 vs 6.06 (ms) 0.55
Pow(1.7,x) [ -30.0, 30.0] 0.003042 % 0.049457 % 2121.500000 6.42 vs 33.01 (ms) 0.19
Pow(x,1.7) [ 0.6, 20.0] 0.006756 % 0.055199 % 0.015030 6.84 vs 29.47 (ms) 0.23
Fmod(x,1.7) [ -10000.0, 10000.0] 0.126278 % 7214.285714 % 1.700552 13.01 vs 26.47 (ms) 0.49
Fmod(1.7,x) [ -10000.0, 10000.0] 0.000381 % 15.789474 % 0.000000 5.76 vs 14.26 (ms) 0.40