Acceleration with long vector architectures: Implementation and evaluation of the FFT kernel on NEC SX‐Aurora and RISC‐V vector extension

Abstract

SummaryNovel architectures leveraging long and variable vector lengths like the NEC SX‐Aurora or the vector extension of RISCV are appearing as promising solutions on the supercomputing market. These architectures often require re‐coding of scientific kernels. For example, traditional implementations of algorithms for computing the fast Fourier transform (FFT) cannot take full advantage of vector architectures. In this article, we present the implementation of FFT algorithms able to leverage these novel architectures. We evaluate these codes on NEC SX‐Aurora , comparing them with the optimized NEC libraries; and in a prototype of a RISC‐V core with a vector processing unit. We present the benefits and limitations of two approaches of RADIX‐2 FFT vector implementations. We show that our approach makes better use of the vector unit of the NEC SX‐Aurora , reaching higher or equal performance than the optimized NEC library. More generally, we prove the importance of maximizing the vector length usage of the algorithm, taking advantage of the FFT properties to reduce long‐latency vector operations, and reordering the instructions according to the specific hardware features to boost the performance of FFT‐like computational kernels.

Ab initio guided atomistic modelling of nanomaterials on exascale high-performance computing platforms

“…Also, modern CPUs with open instruction set specifications, such as RISC-V, have the ability to operate very long vector units (dynamically adapting to up to 16 kbits, i.e. 256 double-precision elements per instruction), being quite promising candidates to build new HPC platforms [34]. All these architectures, along with new accelerators, will play an important role in the so-called post-exascale era, which is expected to start becoming a reality beyond 2025.…”

Enhancing Fault Tolerance in High-Performance Computing: A real hardware case study on a RISC-V Vector Processing Unit

“…The renewed interest in vector processors for HPC and AI has been signified by the advent of various ISA extensions like Intel AVX-512 [10], ARM SVE [11] and RISC-V RVV [12], as well as vector-enabled supercomputers like NEC SX Architecture [13] and Fugaku [14]. Operations in vector processors are executed on arrays of data, called vector registers, composed of vector elements.…”