In this bimonthly feature, HPCwire highlights newly published research in the high-performance computing community and related domains. From parallel programming to exascale to quantum computing, the details are here.
As supercomputers soar in computing ability, the energy necessary to power them is soaring in tandem. These researchers – a team from Nigeria and Korea – present a database to sample system components’ power consumption at regular intervals to serve as input data for energy efficiency optimization. Further, they discuss applications of AI to assess this database and automatically adjust components to maximize energy efficiency.
Authors: Anabi Hilary Kelechi, Mohammed H. Alsharif, Okpe Jonah Bameyi, Paul Joan Ezra, Iorshase Kator Joseph, Aaron-Anthony Atayero, Zong Woo Geem and Junhee Hong.
When volcanoes erupt, they spew massive amounts of sulfur dioxide into the atmosphere – sometimes enough to change local or global climates for weeks or months. A team of researchers from Guangzhou, Beijing and Jülich conducted a high-performance case study of these volcanic emissions, using a particle dispersion model running on the Tianhe-2 supercomputer to estimate volcanic emissions.
Authors: Mingzhao Liu, Yaopeng Huang, Lars Hoffman, Chunyan Huang, Pin Chen and Yi Heng.
As cloud computing continues its upward trajectory, its applications in bioinformatics are growing in number. In this paper, a team of Japanese researchers discuss a port of MEGADOCK, a protein-protein interaction prediction model, to Microsoft Azure. The researchers find a strong scaling value for both the CPU and GPU instances and cite high portability.
Authors: Masahito Ohue, Kento Aoyama and Yutaka Akiyama.
Even as mobile gaming becomes increasingly popular, the types of games playable on mobile systems remains hardware-limited. Cloud gaming – rendering games remotely and streaming them in real-time to client devices – is emerging as a solution to this bottleneck. This paper, written by a team of researchers from Indonesia, discusses the use of HPC cloud environments to manage and distribute these gaming workloads.
Authors: Dedy Prasetya Kristiadi, Ferry Sudarto, Evan Fabian Rahardja, Naufal Rayfi Hafizh, Christopher Samuel and Harco Leslie Hendric Spits Warnars.
Understanding the applications running on a supercomputer is crucial for planning its design, development and operation. In this paper, a team from Argonne National Laboratory, Oak Ridge National Laboratory and Northern Illinois University outline the use of correlative analysis of subsystem logs to show patterns in the applications running on leadership supercomputers.
Authors: Zhengchun Liu, Ryan Lewis, Rajkumar Kettimuthu, Kevin Harms, Philip Carns, Nageswara Rao, Ian Foster and Michael E. Papka.
Climate modeling is typically conducted on supercomputers – but with the boom in cloud computing, new possibilities are emerging. These authors (from Australia, Spain and the UK) discuss the use of cloud computing for climate science by evaluating two different climate models customized to run in public cloud computing environments by Amazon, Google and Microsoft.
Authors: Diego Montes, Juan A. Añel, David C. H. Wallom, Peter Uhe, Pablo V. Caderno and Tomás F. Pena.
With the increasing complexity and heterogeneity of HPC clusters, effective programming for these platforms is becoming a more difficult task. In this paper, four researchers from the Barcelona Supercomputing Center conduct micro-benchmarking of three HPC clusters with diverse architectures present in the Top500 list of the world’s most powerful publicly ranked supercomputers using a production computational fluid dynamics (CFD) code.
Authors: Fabio Banchelli, Marta Garcia-Gasulla, Guillaume Houzeaux and Filippo Mantovani.
Do you know about research that should be included in next month’s list? If so, send us an email at [email protected]. We look forward to hearing from you.