ScalSALE

Re'em Harel

Re'em Harel

Be'er Sheva, South District

Scalable SALE Benchmark Framework for Supercomputers. Utilizing MPI + OpenMP for maximum efficiency (via oneAPI). ...learn more

Project status: Under Development

HPC

Intel Technologies
oneAPI

Docs/PDFs [1]Code Samples [1]

Overview / Usage

Supercomputers worldwide provide the necessary infrastructure for groundbreaking research. As the demand for enhancing their performance constantly increases, new systems are built worldwide to supply this demand at an increased pace, volume, and power. However, most supercomputers are not designed equally due to different desired figure of merit, which is derived from the computational bounds of the targeted scientific applications' portfolio. In turn, the design of such computers becomes a budget-constrained optimization process that strives to achieve the best performances possible in a multi-parameters search space. Therefore, verifying and evaluating whether a supercomputer can achieve its desired goal becomes a tedious and complex task. For this purpose, many full, mini, proxy, and benchmark applications have been introduced in the attempt to represent scientific applications partially. Nevertheless, as these benchmarks are hard to expand, update technologically, and most importantly, are over-simplified compared to scientific applications that tend to couple multiple scientific domains, they fail to represent their true scaling capabilities. We suggest a new physical scalable benchmark, namely ScalSALE, based on the well-known SALE scheme. ScalSALE's main goal is to provide a gold-standard benchmark application that can incorporate multi-physical schemes while maintaining scalable and efficient execution times. By expanding ScalSALE, the gap between the over-simplified benchmarks and scientific applications can be bridged.

Methodology / Approach

Written in Modern Fortran and applying OOP design patterns that will allow the easy expansion of ScalSALE while also maintaining efficient runtime.

One core feature is using MPI and OpenMP parallelization paradigms with the oneAPI (as stated in our paper) to allow an easy-to-use framework, with one simple compiler

Technologies Used

oneAPI, we tested the scaling capabilities on Intel Xeon Gold.

Documents and Presentations

Repository

https://github.com/Scientific-Computing-Lab-NRCN/ScalSALE

Collaborators

1 Result

1 Result

Comments (0)