Developer Reference

Intel® oneAPI Math Kernel Library Vector Mathematics Performance and Accuracy Data

Download PDF
ID 772989
Date 12/04/2020
Public
Document Table of Contents
Document Table of Contents x
Intel® oneAPI Math Kernel Library Vector Mathematics Performance and Accuracy Data
Intel® oneAPI Math Kernel Library Vector Mathematics Performance and Accuracy Data x
Real Functions Complex Functions Notices and Disclaimers
Real Functions x
Performance of All Real VM Functions Measured Accuracy of All Real VM Functions Trigonometric Functions Hyperbolic Functions Power Root Exponential Logarithmic Arithmetic Rounding Special Miscellaneous
Trigonometric Functions x
Acos Acospi Asin Asinpi Atan Atanpi Atan2 Atan2pi Cos Cosd Cospi Sin SinCos Sind Sinpi Tan Tand Tanpi
Hyperbolic Functions x
Acosh Asinh Atanh Cosh Sinh Tanh
Power x
Pow Powr Powx Pow2o3 Pow3o2
Root x
Cbrt Hypot Inv InvCbrt InvSqrt Sqrt
Exponential x
Exp Expm1 Exp2 Exp10
Logarithmic x
Ln Logb Log2 Log10 Log1p
Arithmetic x
Abs Add Div Fmod Frac LinearFrac Performance: Accuracy: Performance Graphs: Mul Remainder Sqr Sub
Rounding x
Ceil Floor Modf NearbyInt Rint Round Trunc
Special x
CdfNorm CdfNormInv Erf Erfc ErfcInv ErfInv ExpInt1 LGamma TGamma
Miscellaneous x
CopySign Fdim Fmax Fmin MaxMag MinMag NextAfter
Complex Functions x
Performance of All Complex VM Functions Measured Accuracy of All Complex VM Functions Trigonometric Functions Hyperbolic Functions Power Root Exponential Logarithmic Arithmetic
Trigonometric Functions x
Acos Asin Atan CIS Cos Sin Tan
Hyperbolic Functions x
Acosh Asinh Atanh Cosh Sinh Tanh
Power x
Pow Powx
Root x
Sqrt
Exponential x
Exp
Logarithmic x
Ln Log10
Arithmetic x
Abs Add Arg Conj Div Mul MulByConj Sub
Intel® oneAPI Math Kernel Library Vector Mathematics Performance and Accuracy Data

LinearFrac

Performs linear fraction transformation of vectors a and b with scalar parameters.

Performance:

  Intel® Xeon® Processor E5-2680 Intel® Xeon® Processor E5-2699 v3 Intel® Xeon® Processor E5-2699 v4 Intel® Xeon® Gold 6148 Processor
single HA N/A N/A N/A N/A
LA N/A N/A N/A N/A
EP 1.28 1.55 1.18 1.05
double HA N/A N/A N/A N/A
LA N/A N/A N/A N/A
EP 11.11 4.68 3.58 1.66
Notes:
  1. Units - CPE (Clocks per Element)

  2. Performance was measured on "working" interval (same for both single and double precision: [ 0.000001, 100000.0] for arguments a, b, scalea, and scaleb; [ 0.000001, 100000.0 ] for arguments shifta, and shiftb)

  3. Data - vectors of 1000 elements with random generated numbers.

  4. All data was collected with 64-bit libraries.

Accuracy:

  Intel® Xeon® Processor E5-2680 Intel® Xeon® Processor E5-2699 v3 Intel® Xeon® Processor E5-2699 v4 Intel® Xeon® Gold 6148 Processor
single HA N/A N/A N/A N/A
LA N/A N/A N/A N/A
EP 80.40 80.40 80.40 80.40
double HA N/A N/A N/A N/A
LA N/A N/A N/A N/A
EP 1.67E+05 1.67E+05 1.67E+05 3.41E+03
NOTE:
Units - ULP (Units in the Last Place).

Performance Graphs:

Parent topic: Arithmetic
Level Two Title