面向处理器微架构设计空间探索的加速方法综述.pdfVIP

计算机研究与发展

Journal

of

Computer

Research

and

Development62（1）：22−57，2025

面向处理器微架构设计空间探索的加速方法综述

1,231,21,21,21,21,2

王铎刘景磊严明玉滕亦涵韩登科叶笑春范东睿

1

（处理器芯片全国重点实验室（中国科学院计算技术研究所）北京100190）

2

（中国科学院大学计算机科学与技术学院北京100049）

3

（中国移动研究院北京100053）

（wangduo@）

AccelerationMethodsforProcessorMicroarchitectureDesignSpaceExploration:A

Survey

1,2

3

1,2

1,2

1,2

1,2

1,2

Wang

Duo,

Liu

Jinglei,

Yan

Mingyu,

Teng

Yihan,

Han

Dengke,

Ye

Xiaochun,

and

Fan

Dongrui

1

（StateKeyLabofProcessors(InstituteofComputingTechnology,ChineseAcademyofSciences),Beijing100190）

2

（SchoolofComputerScienceandTechnology,UniversityofChineseAcademyofSciences,Beijing100049）

3

（ChinaMobileCommunicationsResearchInstitute,Beijing100053）

AbstractCentral

processing

unit

is

the

most

important

computing

infrastructure

nowadays.

To

maximize

the

profit,

architects

design

the

processor

microarchitecture

by

trading-off

multiple

objectives

including

performance,

power,

and

area.

However,

because

of

the

tremendous

instructions

of

workloads

running

on

the

processors,

the

evaluation

of

individual

microarchitecture

design

point

costs

minutes

to

hours.

Furthermore,

the

design

space

of

the

microar-

chitecture

is

huge,

which

results

that

the

exploration

of

comprehensive

design

space

is

unrealistic.

Therefore,

many

machine-learning-assisted

design

space

exploration

acceleration

methods

are

proposed

to

reduce

the

size

of

evaluated

desig