Apex: the apl parallel executor
Apex: the apl parallel executor
复制标题
Apex：apl并行执行器
DOI：
--
复制
发表时间：
1997
期刊：
影响因子：
--
通讯作者：
Robert Bernecky
中科院分区：
文献类型：
--
作者：
Robert Bernecky
APEX：APL平行执行者Robert Bernecky科学系计算机科学大学多伦多1997年Apex是APL-to-sisal编译器，产生了高性能的ANCE，便携式，并行代码，其执行速度比解释快几百倍APL，与Fortran竞争的内核连续性能。初步结果表明，可接受的多处理器加速是可以实现的。 Apex生成的代码的出色运行时间性能源于对程序执行的各个方面的关注：消除运行时语法分析，降低了STUP成本，代数标识和短语识别检测特殊情况，某些矩阵PR ODUCTS ODUCTS利用A Sparsematrix代数的概括，环融合和复制优化消除了许多阵列值值的临时。此外，编译器利用静态单分配和RRAR形态，我们将数据流分析对数组的概括，以生成使用卓越算法和更简单的存储类型的运行时间原始s。 APL的扩展，包括排名，削减和减少二元的单调操作数，改善编译和解释的代码性能。
APEX: the APL Parallel Executor Robert Bernecky Master of Science Graduate Department of Computer Science University of Toronto 1997 APEX is an APL-to-SISAL compiler, generating high-perform ance, portable, parallel code that executes up to several hundred times faster than interpreted APL, wit h serial performance of kernels competitive with FORTRAN. Preliminary results indicate that acceptabl e multi-processor speedup is achievable. The excellent run-time performance of APEX-generated code arises from attention to all aspects of program execution: run-time syntax analysis is eliminated , s tup costs are reduced, algebraic identities and phrase recognition detect special cases, some matrix pr oducts exploit a generalization of sparsematrix algebra, and loop fusion and copy optimizations elim inate many array-valued temporaries. In addition, the compiler exploits Static Single Assignment a nd rray morphology, our generalization of data flow analysis to arrays, to generate run-time primitive s that use superior algorithms and simpler storage types. Extensions to APL, including rank, cut, and a monadic operand for dyadic reduction, improve compiled and interpreted code performance.