Architectural Specialization for Inter-Iteration Loop Dependence Patterns
MICRO(2014)
摘要
Hardware specialization is an increasingly common technique to enable improved performance and energy efficiency in spite of the diminished benefits of technology scaling. This paper proposes a new approach called explicit loop specialization (XLOOPS) based on the idea of elegantly encoding inter-iteration loop dependence patterns in the instruction set. XLOOPS supports a variety of inter-iteration data-and control-dependence patterns for both single and nested loops. The XLOOPS hardware/software abstraction requires only lightweight changes to a general-purpose compiler to generate XLOOPS binaries and enables executing these binaries on: (1) traditional micro architectures with minimal performance impact, (2) specialized micro architectures to improve performance and/or energy efficiency, and (3) adaptive micro architectures that can seamlessly migrate loops between traditional and specialized execution to dynamically trade-off performance vs. Energy efficiency. We evaluate XLOOPS using a vertically integrated research methodology and show compelling performance and energy efficiency improvements compared to both simple and complex general-purpose processors.
更多查看译文
关键词
hardware specialization,instruction set,traditional microarchitectures,adaptive microarchitectures,specialized microarchitectures,general-purpose compiler,complex general-purpose processors,energy conservation,architectural specialization,interiteration data-and control-dependence patterns,single loops,xloops binaries,vertically integrated research methodology,simple general-purpose processors,xloops hardware-software abstraction,instruction sets,program control structures,interiteration loop dependence patterns,energy efficiency,explicit loop specialization,nested loops,program compilers,datacenter,compiler,kernel,microarchitecture,optimization,hardware,cache,out of order,resource sharing,registers
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络