Floating point operation

from Wikipedia, the free encyclopedia

Floating-point operations are called commands and calculations in computer programs , the floating-point ( English floating point numbers) use. These operations are required in order to be able to calculate with fractional numbers or numbers in larger number ranges. An arithmetic coprocessor (floating point processor or floating point unit) is able to carry out floating point operations directly. If this is not available, these operations have to be translated into arithmetic operations with integers , for example by a compiler .

Floating Point Operations Per Second (short FLOPS; English for floating point operations per second) is a unit of measurement for the speed of computer systems or processors and describes the number of floating point number operations (additions or multiplications) that can be carried out by them per second.

literature

  • Ulrich Kulisch : The fifth floating point operation for top performance computers or the accumulation of floating point numbers and products in fixed point arithmetic . Institute for Applied Mathematics, Karlsruhe 1997.