GRNTI 50.07 Теоретические основы вычислительной техники
BBK 3297 Вычислительная техника
In the paper we present the situation which had required visualization of a large amount of non-trivial objects, such as supercomputer’s tasks. The method of visualization of these objects was hard to find. Then we used additional information about an extra structure on those objects. This knowledge led us to an idea of grouping the objects into new generalized ones. Those new artificial objects were easy to visualize due to their small quantity. And they happened to be enough for the cognition of the original problem. That was a successful change of point of view. As a whole, our work belongs to a high-performance computing performance visualization area. It gains valuable attention from scientists over the whole world, for example [1-2].
high-performance computing, software visualization, projection, multi-dimensional visualization
1. Showerman M. Real Time Visualization of MonitoringData for Large Scale HPC Systems // 2015 IEEEInternational Conference on Cluster Computing. IEEE,2015. Pp. 706-709. doi.org/10.1109/CLUSTER.2015.122
2. Nikitenko, D., Zhumatiy, S., Shvets, P.: Making large-scalesystems observable – another inescapable step towardsexascale. Supercomput. Front. Innov. J. 3(2), 72–79 (2016).doi.org/10.14529/jsfi160205
3. Igumnov A. S., Bersenev A. Y., Popel A. A. and Vasev P.A. Studying the Latency of Concurrent Access to NetworkStorage // In proceedings of 13th international conferenceon Parallel computational technologies (PCT’2019),Kaliningrad, Russia, 2019. Pp. 66-77.elibrary.ru/item.asp?id=37327250