Performance Left on the Table: An Evaluation of Compiler Autovectorization for RISC-V

Performance Left on the Table: An Evaluation of Compiler Autovectorization for RISC-V
复制标题

剩下的性能:RISC-V 编译器自动向量化的评估

DOI:
10.1109/mm.2022.3184867
复制
发表时间:
2022
期刊:
影响因子:
3.6
通讯作者:
Sampson, Adrian
Sampson, Adrian
中科院分区:
计算机科学3区
文献类型:
--
作者:
Adit, Neil;Sampson, Adrian

文献摘要

参考文献

被引文献

相似文献

下一代长度不可知的向量指令集架构(ISA)设计、RISC-V向量扩展和ARM的可扩展向量扩展使软件可跨采用不同向量引擎的硬件实现进行移植。虽然传统的定长单指令多数据ISA指令(如Intel AVX和ARM Neon)享有成熟的编译器自动向量化支持,但对于这些与长度无关的ISA,编译器支持仍在不断涌现。这项工作研究了编译器的缺陷,这些缺陷构成了长度不可知体系结构和固定长度体系结构之间自动向量化能力的差距。我们研究了LLVM对RISC-V向量扩展和传统向量ISA的支持。我们研究了一组合成标量循环来比较两种设置下的支持广度,并检查了一个真实的基准测试套件来比较自动向量化和手动向量化的RISC-V代码。我们使用这两项研究来提炼出一组建议,用于工程改进和未来对长度不可知向量编程的编译器和编程模型的研究。
Next-generation length-agnostic vector instruction set architecture (ISA) designs, the RISC-V vector extension, and ARM’s scalable vector extension enable software portability across hardware implementations with different vector engines. While traditional, fixed-length single-instruction–multiple-data ISA instructions, such as Intel AVX and ARM Neon, enjoy mature compiler support for automatic vectorization, compiler support is still emerging for these length-agnostic ISAs. This work studies the compiler shortcomings that constitute the gap in autovectorization capabilities between length-agnostic and fixed-length architectures. We examine LLVM’s support for both the RISC-V vector extension and traditional vector ISAs. We study a set of synthetic scalar loops to compare the breadth of support in the two settings, and we examine a real benchmark suite to compare autovectorized to hand-vectorized RISC-V code. We use both studies to distill a set of recommendations for engineering improvements and future research in compilers and programming models for length-agnostic vector programming.
通过客观提取有用信息来评估自动矢量化编译器
DOI: --
发表时间: 2019
期刊: ACM Transactions on Architecture and Code Optimization (TACO)
影响因子: --
作者:
Sergi Siso;W. Armour;Jeyarajan Thiyagalingam
通讯作者: Jeyarajan Thiyagalingam
通过推测向量化过去的依赖分支
DOI: --
发表时间: 2013
期刊: Proceedings of the 22nd International Conference on Parallel Architectures and Compilation Techniques
影响因子: --
作者:
Majedul Haque Sujon;R. C. Whaley;Qing Yi
通讯作者: Qing Yi
矢量长度不可知代码的性能分析
DOI: --
发表时间: 2019
期刊: International Symposium on High Performance Computing Systems and Applications
影响因子: --
作者:
Angela Pohl;Mirko Greese;Biagio Cosenza;B. Juurlink
通讯作者: B. Juurlink
Intel 64 和 Intel Xeon Phi 上循环的自动矢量化:分析和评估
DOI: --
发表时间: 2017
期刊: International Conference on Parallel Architectures and Compilation Techniques
影响因子: --
作者:
O. Moldovanova;M. Kurnosov
通讯作者: M. Kurnosov
用于设计和评估矢量架构的 RISC-V 模拟器和基准套件
DOI: 10.1145/3422667
发表时间: 2020
期刊: ACM Transactions on Architecture and Code Optimization (TACO)
影响因子: --
作者:
Cristóbal Ramírez;César;Oscar Palomar;O. Unsal;M. A. Ramírez;A. Cristal
通讯作者: A. Cristal