基于CPU微架构人工智能任务性能优化的策略研究  

Research on Strategy for Performance Optimization of Artificial Intelligence Tasks Based on CPU Microarchitecture

作　　者：王逸之 WANG Yizhi(University of Alberta,Alberta Canada)

机构地区：[1]阿尔伯塔大学,加拿大阿尔伯塔

出　　处：《软件》2024年第9期121-123,共3页Software

摘　　要：本文针对人工智能相关任务的需求,提出对CPU微架构的优化方案。通过对CPU流水线的优化,避免了分支预测可能失败的情况,提高了计算人工智能相关任务中需CPU执行部分的执行效率。This article proposes an optimization plan for CPU microarchitecture to meet the requirements of artificial intelligence related tasks.By optimizing the CPU pipeline,the possibility of branch prediction failure has been avoided,and the execution efficiency of CPU required parts in artificial intelligence related tasks has been improved.

关 键 词：CPU微架构 人工智能 性能优化 

分 类 号：TP333[自动化与计算机技术—计算机系统结构]