A new open-source SIMD vector libm fully implemented with high-level scalar C

Lauter, Christoph

doi:10.1109/acssc.2016.7869070

Cited by 13 publications

(15 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, we could concentrate on the design of efficient exponential or trigonometric functions, which are also widely used in HPC. Second we could adapt our work to produce programs with different accuracy levels, like in [5] where 8 ulp is guaranteed.…”

Section: Discussionmentioning

confidence: 99%

“…The VDT is another open source effort [4]: in terms of accuracy, log(x) function is announced with at most 2 bits different compared to AMD LibM. Two recent projects are Yeppp 4 and SLEEF, 5 the latter providing 3-ulp vectorized function versions. Finally in 2016, a new SIMD vector library is proposed [5].…”

Section: Meta-implementation Of Vectorized Logarithm Function In Binamentioning

confidence: 99%

“…Two recent projects are Yeppp 4 and SLEEF, 5 the latter providing 3-ulp vectorized function versions. Finally in 2016, a new SIMD vector library is proposed [5]. Written in high-level C language, log(x) is accurate to at most 1.0825 ulp.…”

Section: Meta-implementation Of Vectorized Logarithm Function In Binamentioning

confidence: 99%

See 2 more Smart Citations

Meta-implementation of vectorized logarithm function in binary floating-point arithmetic

Saint-Genies

Brunie

Revy

2018

2018 IEEE 29th International Conference on Application-Specific Systems, Architectures and Processors (ASAP)

View full text Add to dashboard Cite

Besides scalar instructions, modern microarchitectures also provide support for vector instructions. They enable to treat packed inputs (typically 4 or 8) in a single instruction. The challenge is now to write vector programs to support mathematical functions like sin, cos, exp, log, • • • which efficiently exploit those vector instructions. This article focuses on the design of vectorized implementation of log(x) function, and more particularly on its automation for different formats and micro-architectures. First it rewrites a classic range reduction in a branchless fashion so as to use at best recent micro-architecture features, like rcp (reciprocal) instruction, and to treat all inputs in the same flow. Second it details rigorously how to achieve "faithfully rounded" implementations. Third it shows how to automate this implementation process using the MetaLibm framework, on SSE/AVX and AVX2 supporting micro-architectures. Finally we illustrate that this process enables to achieve high throughput implementations for the binary32 and binary64 formats in a fully automated way.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Meta-implementation Of Vectorized Logarithm Function In Binamentioning

confidence: 99%

See 1 more Smart Citation

Meta-implementation of vectorized logarithm function in binary floating-point arithmetic

Saint-Genies

Brunie

Revy

2018

2018 IEEE 29th International Conference on Application-Specific Systems, Architectures and Processors (ASAP)

View full text Add to dashboard Cite

show abstract

“…In the main loop, the algorithm finds the remainder of n/d, and at line 22, the correct sign is assigned to the resulting remainder. 5. https://sleef.org/ Product of DD number x and DP number y, where y = 2 N ddnormalize d2 d2(x)…”

Section: Appendix a Annotated Source Code Of Tanmentioning

confidence: 99%

SLEEF: A Portable Vectorized Library of C Standard Mathematical Functions

Shibata

Petrogalli

2020

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

In this paper, we present techniques used to implement our portable vectorized library of C standard mathematical functions written entirely in C language. In order to make the library portable while maintaining good performance, intrinsic functions of vector extensions are abstracted by inline functions or preprocessor macros. We implemented the functions so that they can use sub-features of vector extensions such as fused multiply-add, mask registers and extraction of mantissa. In order to make computation with SIMD instructions efficient, the library only uses a small number of conditional branches, and all the computation paths are vectorized. We devised a variation of the Payne-Hanek argument reduction for trigonometric functions and a floating point remainder, both of which are suitable for vector computation. We compare the performance of our library to Intel SVML. Index Terms-Parallel and vector implementations, SIMD processors, elementary functions, floating-point arithmetic ✦ 1. https://unity3d.com/.

show abstract

“…For this purpose, we have relied on external vectorized mathematical libraries. Among the various available, we have tested the SIMD vector libm [6] and the CERN's VDT [7] library. The main advantage of these two libraries with respect to others (i.e.…”

Section: Manual Optimization Through Vectorizationmentioning

confidence: 99%

Performance optimization of the air shower simulation program for the Cherenkov Telescope Array

et al. 2019

View full text Add to dashboard Cite

The Cherenkov Telescope Array (CTA), currently under construction, is the next-generation instrument in the field of very high energy gammaray astronomy. The first data are expected by the end of 2018, while the scientific operations will start in 2022 for a duration of about 30 years. In order to characterize the instrument response to the Cherenkov light emitted when cosmic ray showers develop in the atmosphere, detailed Monte Carlo simulations will be regularly performed in parallel to CTA operation. The estimated CPU time associated to these simulations is very high, of the order of 200 millions HS06 hours per year. Reducing the CPU time devoted to simulations would allow either to reduce infrastructure cost or to better cover the large phase space. In this paper, we focus on the main computing step (70% of the whole CPU time) implemented in the CORSIKA program, and specifically on the module responsible for the propagation of Cherenkov photons in the atmosphere. We present our preliminary studies about different options of code optimization, with a particular focus on vectorization facilities (SIMD instructions). Our proposals take care, as automatically as possible, of the hardware portability constraints introduced by the grid computing environment that hosts these simulations. Performance evaluation in terms of running-time and accuracy is provided.

show abstract

A new open-source SIMD vector libm fully implemented with high-level scalar C

Cited by 13 publications

References 11 publications

Meta-implementation of vectorized logarithm function in binary floating-point arithmetic

Meta-implementation of vectorized logarithm function in binary floating-point arithmetic

SLEEF: A Portable Vectorized Library of C Standard Mathematical Functions

Performance optimization of the air shower simulation program for the Cherenkov Telescope Array

Contact Info

Product

Resources

About