Add to ORKGThe definitive version of this preliminary work is available here https://hal.archives-ouvertes.fr/hal-01236919v2 or here https://doi.org/10.1002/cpe.3953.We present the SyHD software developed to perform source-to-source transformation improving accuracy without impacting execution time too much. SyHD synthesizes C source code for both accuracy and execution-time criteria. It uses compensation for improving accuracy and transformation strategies for reducing the impact of this improvement on execution time
International audienceIn embedded systems, many numerical algorithms are implemented with fixed-poin...
International audienceFPGAs are well known for their ability to perform non-standard computations no...
Approximating ideal program outputs is a common technique for solving computationally difficult prob...
The definitive version of this preliminary work is available here https://hal.archives-ouvertes.fr/h...
International audienceNumerical programs with IEEE 754 floating-point computations may suffer from i...
International audienceNumerical programs with IEEE 754 floating-point computations may suffer from i...
Numerical accuracy and execution time of programs using the floating-point arithmetic are major chal...
International audienceHigh-level synthesis (HLS) is a big step forward in terms of design productivi...
The floating-point numbers used in computer programs are a finite approximation of real numbers. In ...
In high performance computing, nearly all the implementations and published experiments use floatin...
Over the last decade, guaranteeing the accuracy of computations relying on the IEEE754 floating-poin...
International audienceError-tolerating applications are increasingly common in the emerging field of...
<p>In this thesis, we design frameworks for efficient and accurate floating point computation. The p...
Many scenarios demand a high processing power often combined with a limited energy budget. A way to ...
While tremendously useful, automated techniques for tuning the precision of floating-point programs ...
International audienceIn embedded systems, many numerical algorithms are implemented with fixed-poin...
International audienceFPGAs are well known for their ability to perform non-standard computations no...
Approximating ideal program outputs is a common technique for solving computationally difficult prob...
The definitive version of this preliminary work is available here https://hal.archives-ouvertes.fr/h...
International audienceNumerical programs with IEEE 754 floating-point computations may suffer from i...
International audienceNumerical programs with IEEE 754 floating-point computations may suffer from i...
Numerical accuracy and execution time of programs using the floating-point arithmetic are major chal...
International audienceHigh-level synthesis (HLS) is a big step forward in terms of design productivi...
The floating-point numbers used in computer programs are a finite approximation of real numbers. In ...
In high performance computing, nearly all the implementations and published experiments use floatin...
Over the last decade, guaranteeing the accuracy of computations relying on the IEEE754 floating-poin...
International audienceError-tolerating applications are increasingly common in the emerging field of...
<p>In this thesis, we design frameworks for efficient and accurate floating point computation. The p...
Many scenarios demand a high processing power often combined with a limited energy budget. A way to ...
While tremendously useful, automated techniques for tuning the precision of floating-point programs ...
International audienceIn embedded systems, many numerical algorithms are implemented with fixed-poin...
International audienceFPGAs are well known for their ability to perform non-standard computations no...
Approximating ideal program outputs is a common technique for solving computationally difficult prob...