English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

In this paper, we analyze the performance and energy consumption of anArm-based high-performance computing (HPC) system developed within the Europeanproject Mont-Blanc 3. This system, called Dibona, has been integrated byATOS/Bull, and it is powered by the latest Marvell's CPU, ThunderX2. This CPUis the same one that powers the Astra supercomputer, the first Arm-basedsupercomputer entering the Top500 in November 2018. We study frommicro-benchmarks up to large production codes. We include an interdisciplinaryevaluation of three scientific applications (a finite-element fluid dynamicscode, a smoothed particle hydrodynamics code, and a lattice Boltzmann code) andthe Graph 500 benchmark, focusing on parallel and energy efficiency as well asstudying their scalability up to thousands of Armv8 cores. For comparison, werun the same tests on state-of-the-art x86 nodes included in Dibona and theTier-0 supercomputer MareNostrum4. Our experiments show that the ThunderX2 hasa 25% lower performance on average, mainly due to its small vector unit yetsomewhat compensated by its 30% wider links between the CPU and the mainmemory. We found that the software ecosystem of the Armv8 architecture iscomparable to the one available for Intel. Our results also show that ThunderX2delivers similar or better energy-to-solution and scalability, proving thatArm-based chips are legitimate contenders in the market of next-generation HPCsystems.

Performance and energy consumption of HPC workloads on a cluster based on Arm ThunderX2 CPU