Achieving High Performance the Functional Way: Expressing High-Performance Optimizations as Rewrite Strategies.
Optimizing programs to run efficiently on modern parallel hardware is hard but crucial for many applications. The predominantly used imperative languages force the programmer to intertwine the code describing functionality and optimizations. This results in a portability nightmare that is particular...
| Publicado en: | Communications of the ACM Vol. 66; no. 3; pp. 89 - 98 |
|---|---|
| Autores principales: | , , , , , |
| Formato: | Artículo |
| Publicado: |
Association for Computing Machinery
Mar2023
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=162018630&site=ehost-live header: @attributes: shortDbName: hlh uiTerm: 162018630 longDbName: Humanities International Complete uiTag: AN controlInfo: bkinfo: jinfo: jid: 00010782 ACM jtl: Communications of the ACM issn: 00010782 maglogo: N pubinfo: dt: Mar2023 vid: 66 iid: 3 pid: 68 pub: Association for Computing Machinery artinfo: ui: 162018630 10.1145/3580371 ppf: 89 ppct: 9 formats: tig: atl: Achieving High Performance the Functional Way: Expressing High-Performance Optimizations as Rewrite Strategies. aug: au: Hagedorn, Bastian Lenfers, Johannes Koehler, Thomas Xueying Qin Gorlatch, Sergei Steuwer, Michel affil: NVIDIA, Würselen, Germany University of Münster, Germany University of Glasgow, U.K. The University of Edinburgh, U.K. su: OPL (Computer program language) Programming languages Domain-specific programming languages Computer programming Electronic data processing Computer software development sug: subj: OPL (Computer program language) Programming languages Domain-specific programming languages Computer programming Electronic data processing Computer software development ab: Optimizing programs to run efficiently on modern parallel hardware is hard but crucial for many applications. The predominantly used imperative languages force the programmer to intertwine the code describing functionality and optimizations. This results in a portability nightmare that is particularly problematic given the accelerating trend toward specialized hardware devices to further increase efficiency. Many emerging domain-specific languages (DSLs) used in performance-demanding domains such as deep learning attempt to simplify or even fully automate the optimization process. Using a high-level--often functional--language, programmers focus on describing functionality in a declarative way. In some systems such as Halide or TVM, a separate schedule specifies how the program should be optimized. Unfortunately, these schedules are not written in well-defined programming languages. Instead, they are implemented as a set of ad hoc predefined APIs that the compiler writers have exposed. In this paper, we show how to employ functional programming techniques to solve this challenge with elegance. We present two functional languages that work together--each addressing a separate concern. RISE is a functional language for expressing computations using well-known data-parallel patterns. ELEVATE is a functional language for describing optimization strategies. A high-level RISE program is transformed into a low-level form using optimization strategies written in ELEVATE. From the rewritten low-level program, high-performance parallel code is automatically generated. In contrast to existing high-performance domain-specific systems with scheduling APIs, in our approach programmers are not restricted to a set of built-in operations and optimizations but freely define their own computational patterns in RISE and optimization strategies in ELEVATE in a composable and reusable way. We show how our holistic functional approach achieves competitive performance with the state-of-the-art imperative systems such as Halide and TVM. pubtype: Periodical doctype: Article src: R language: English refInfo: copyright: @attributes: flag: Y dt: @attributes: year: 2023 holdings: @attributes: islocal: N |
|---|