4.7 Article

Kokkos 3: Programming Model Extensions for the Exascale Era

期刊

出版社

IEEE COMPUTER SOC
DOI: 10.1109/TPDS.2021.3097283

关键词

Programming; Hardware; Kernel; Graphics processing units; Layout; Laboratories; Benchmark testing; Performance portability; programming models; high-performance computing; heterogeneous computing; exascale

资金

  1. U.S. Department of Energy's National Nuclear Security Administration [DE-NA-0003525]
  2. U.S. Department of Energy (DOE) [DE-AC05-00OR22725]
  3. Exascale Computing Project [17-SC-20-SC]
  4. DOE Office of Science User Facility [DE-AC02-06CH11357]

向作者/读者索取更多资源

This article introduces the Kokkos Performance Portable Programming Model, which allows developers to write single source applications for diverse high-performance computing architectures. Kokkos provides key abstractions for the compute and memory hierarchy of modern hardware and demonstrates the performance of new features for exascale-era architectures.
As the push towards exascale hardware has increased the diversity of system architectures, performance portability has become a critical aspect for scientific software. We describe the Kokkos Performance Portable Programming Model that allows developers to write single source applications for diverse high-performance computing architectures. Kokkos provides key abstractions for both the compute and memory hierarchy of modern hardware. We describe the novel abstractions that have been added to Kokkos version 3 such as hierarchical parallelism, containers, task graphs, and arbitrary-sized atomic operations to prepare for exascale era architectures. We demonstrate the performance of these new features with reproducible benchmarks on CPUs and GPUs.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据