The aim of this study was to compare the accuracy of the 32-bit IEEE 754 floating point numbers and 32-bit Posit numbers. In order to achieve this goal, two benchmarks were developed in C++. These benchmarks ran a series of matrix multiplications with IEEE 754 floats, posits and 64-bit doubles and computed an error for each by taking the average relative error of all elements in the result matrix. These relative errors were computed by comparing the element from each 32-bit float matrix and the corresponding element in the 64-bit double element matrix. By plotting the results it could be concluded that 32-bit posits produced results with a higher accuracy than 32-bit IEEE 754 floats for small values as elements. It was also clear from the r...
this paper. In 1985, the Institute of Electrical and Electronic Engineers published IEEE Standard 75...
The current IEEE-754 floating point standard was adopted 23 years ago. IEEE chartered a committee to...
The goal of this thesis is to describe factors which influence the accuracy of mathematical calculat...
Posit is a new way of representing floating points in computers. This thesis investigates the precis...
Choosing the correct floating point representation can greatly impact the performance and precision ...
IEEE 754 floating-points are the current standard way to represent real values in computers, but the...
IEEE-754 has been the standard floating-point format for several decades. However, in recent years m...
Posit is a new floating point precision representation that has been developed as an alternative to ...
The posit is a floating-point number format as proposed by John Gustafson to improve upon the accura...
Projecte realitzat en el marc d'un programa de mobilitat amb la Slovenská Technická Univerzita v Bra...
The posit number system is an elegant encoding of floating-point values proposed as a drop-in replac...
Scientific applications rely heavily on floating-point arithmetic and, therefore, are affected by th...
This thesis presents a direct iteration and implementation on a high per-formance architecture for I...
National audienceThe posit number system is proposed as a replacement of IEEE floats. It encodes flo...
The floating point standard IEEE 754 is widely implemented, but many of its capabilities are not wel...
this paper. In 1985, the Institute of Electrical and Electronic Engineers published IEEE Standard 75...
The current IEEE-754 floating point standard was adopted 23 years ago. IEEE chartered a committee to...
The goal of this thesis is to describe factors which influence the accuracy of mathematical calculat...
Posit is a new way of representing floating points in computers. This thesis investigates the precis...
Choosing the correct floating point representation can greatly impact the performance and precision ...
IEEE 754 floating-points are the current standard way to represent real values in computers, but the...
IEEE-754 has been the standard floating-point format for several decades. However, in recent years m...
Posit is a new floating point precision representation that has been developed as an alternative to ...
The posit is a floating-point number format as proposed by John Gustafson to improve upon the accura...
Projecte realitzat en el marc d'un programa de mobilitat amb la Slovenská Technická Univerzita v Bra...
The posit number system is an elegant encoding of floating-point values proposed as a drop-in replac...
Scientific applications rely heavily on floating-point arithmetic and, therefore, are affected by th...
This thesis presents a direct iteration and implementation on a high per-formance architecture for I...
National audienceThe posit number system is proposed as a replacement of IEEE floats. It encodes flo...
The floating point standard IEEE 754 is widely implemented, but many of its capabilities are not wel...
this paper. In 1985, the Institute of Electrical and Electronic Engineers published IEEE Standard 75...
The current IEEE-754 floating point standard was adopted 23 years ago. IEEE chartered a committee to...
The goal of this thesis is to describe factors which influence the accuracy of mathematical calculat...