Optimize for Intel® AVX-512 with or without Intel® AVX-512 hardware

Overview

Optimizing vectorization becomes increasingly important as new technology like Intel® Advanced Vector Extensions 512 (Intel® AVX-512) becomes available. Software must be both efficiently threaded and efficiently vectorized to fully utilize modern hardware.

Get an overview of what changes with Intel AVX-512 and learn how Intel® Advisor 2017 can help you optimize for both Intel AVX-512 and older instruction sets both with and without access to the latest hardware. See how to get accurate repeatable FLOPS metrics, measure the loop footprint to see if it fits into cache, detect unneeded gather/scatters that reduce performance and much more.

[block]pre-patch-disclaimer[/block]

[block]ftc-disclaimer[/block]

产品和性能信息

1

英特尔的编译器针对非英特尔微处理器的优化程度可能与英特尔微处理器相同(或不同)。这些优化包括 SSE2、SSE3 和 SSSE3 指令集和其他优化。对于在非英特尔制造的微处理器上进行的优化,英特尔不对相应的可用性、功能或有效性提供担保。该产品中依赖于微处理器的优化仅适用于英特尔微处理器。某些非特定于英特尔微架构的优化保留用于英特尔微处理器。关于此通知涵盖的特定指令集的更多信息,请参阅适用产品的用户指南和参考指南。

通知版本 #20110804