Efficient Evaluation Methods of Elementary Functions Suitable for SIMD Computation

12 years 5 months ago

Download www.slideshare.net

Data-parallel architectures like SIMD (Single Instruction Multiple Data) or SIMT (Single Instruction Multiple Thread) have been adopted in many recent CPU and GPU architectures. Although some SIMD and SIMT instruction sets include double-precision arithmetic and bitwise operations, there are no instructions dedicated to evaluating elementary functions like trigonometric functions in double precision. Thus, these functions have to be evaluated one by one using an FPU or using a software library. However, traditional algorithms for evaluating these elementary functions involve heavy use of conditional branches and/or table look-ups, which are not suitable for SIMD computation. In this paper, efficient methods are proposed for evaluating the sine, cosine, arc tangent, exponential and logarithmic functions in double precision without table look-ups, scattering from, or gathering into SIMD registers, or conditional branches. We implemented these methods using the Intel SSE2 instruction set ...

Real-time Traffic

Elementary Functions | SIMD |

posted by naokishibata

Post Info
More Details (n/a)

Added	17 Jun 2012
Updated	25 Aug 2012
Authors	Naoki Shibata

Comments (0)

	Video Ads Dissemination through WiFi-Cellular Hybrid Networks 453 views
	Framework for virtual collaboration emphasized by awareness information and asynchronous interaction 439 views
	General Architecture for Hardware Implementation of Genetic Algorithm 404 views
	Cheat detection for MMORPG on P2P environments 489 views
	Distributed market broker architecture for resource aggregation in grid computing environments 406 views

Sciweavers

Efficient Evaluation Methods of Elementary Functions Suitable for SIMD Computation

Elementary Functions | SIMD |

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers