避免 AVX-SSE 转换造成的性能损失

避免 AVX-SSE 转换造成的性能损失 (PDF 678 KB)

Autor: Patrick Konsor (Intel) Zuletzt aktualisiert am 12.02.2016 - 14:08

Ticker Tape:具有风力和空气阻力效果的可扩展 3D 粒子系统

本文将介绍我们模拟 3D 粒子行为的方法,以及我们使用的模块化设计,以达到促进重复使用和实验的目的。此外,我们还会讨论通过线程化以及使用英特尔® SIMD 流指令扩展(英特尔® SSE)实现的性能提升。
Autor: Mike Yi (Intel) Zuletzt aktualisiert am 12.02.2016 - 10:13

SIMD 编程的优势 --TickerTape Part 2

Ticker Tape 是一种技术演示,旨在鼓励开发人员在粒子系统中执行更为复杂的操作。参与该演示的开发人员会运用大量技术,来提高包括多线程和针对英特尔® SIMD 流指令扩展(SSE)的优化等在内的性能
Autor: Administrator Zuletzt aktualisiert am 12.02.2016 - 10:09

在不编写 AVX 代码的情况下使用 AVX

Using AVX Without Writing AVX Code (PDF 260KB)

Autor: Administrator Zuletzt aktualisiert am 11.02.2016 - 14:47

利用英特尔® SIMD 流指令扩展和英特尔® 高级矢量扩展指令集的图像处理加速技术

This article details optimized implementations of data transformations and algorithms together with analysis comparing performance and providing speedup measurements for Intel® SSE optimized code and estimates for Intel® AVX optimized code.
Autor: Petter Larsson (Intel) Zuletzt aktualisiert am 11.02.2016 - 14:34
Blog post

Explore Intel® AVX-512 Code Paths with Intel® Advisor XE while not Having Compatible Hardware

Many factors that can make programs difficult for automatic vectorization.

Autor: Vadim Kartoshkin (Intel) Zuletzt aktualisiert am 11.02.2016 - 09:50
Blog post

Use which hardware PMU events to calculate FLOPS on Intel(R) Xeon Phi(TM) coprocessor?

FLOPS means total floating point operations per second, which is used in High Performance Computing. In general, Intel(R) VTune(TM) Amplifier XE

Autor: Peter Wang Zuletzt aktualisiert am 10.02.2016 - 20:40
Forum topic

SGX and SGX1 of CPUID with SKL emulation

I am confused by CPUID data (see below) of SKL emulation with the latest version (7.39-win) of Intel SDE.

Autor: Jeremy W. Zuletzt aktualisiert am 10.02.2016 - 09:43

Reference Implementations for Intel® Architecture Approximation Instructions VRCP14, VRSQRT14, VRCP28, VRSQRT28, and VEXP2

We are providing source files containing reference implementations for the scalar versions of 10 approximation instructions introduced in the "Intel® Architecture Instruction Set Extensions Programming Reference" document
Autor: Administrator Zuletzt aktualisiert am 10.02.2016 - 07:50
Forum topic

Executed instruction not valid for specified chip (PENTIUM4)

I encounter the following error message with the latest version (7.39-win) of Intel SDE, when I attempt the "-p4" switch. What is the preferred way of using the "-p4" switch?

Autor: Jeremy W. Zuletzt aktualisiert am 10.02.2016 - 07:34
Einzelheiten zur Compiler-Optimierung finden Sie in unserem Optimierungshinweis.