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/

Global: Zendy Tools annual

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

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

While experiments on fusion plasmas produce high-dimensional data time serieswith ever increasing magnitude and velocity, data analysis has been laggingbehind this development. For example, many data analysis tasks are oftenperformed in a manual, ad-hoc manner some time after an experiment. In thisarticle we introduce the DELTA framework that facilitates near real-timestreaming analysis of big and fast fusion data. By streaming measurement datafrom fusion experiments to a high-performance compute center, DELTA allows toperform demanding data analysis tasks in between plasma pulses. This articledescribe the modular and expandable software architecture of DELTA and presentsperformance benchmarks of its individual components as well as of entireworkflows. Our focus is on the streaming analysis of ECEi data measured atKSTAR on NERSCs supercomputers and we routinely achieve data transfer rates ofabout 500 Megabyte per second. We show that a demanding turbulence analysisworkload can be distributed among multiple GPUs and executes in under 5minutes. We further discuss how DELTA uses modern database systems andcontainer orchestration services to provide web-based real-time datavisualization. For the case of ECEi data we demonstrate how data visualizationscan be augmented with outputs from machine learning models. By providingsession leaders and physics operators results of higher order data analysisusing live visualization they may monitor the evolution of a long-pulsedischarge in near real-time and may make more informed decision on how toconfigure the machine for the next shot.

Near real-time streaming analysis of big fusion data