TY - GEN
T1 - Comparing the efficiency of in situ visualization paradigms at scale
AU - Kress, James
AU - Larsen, Matthew
AU - Choi, Jong
AU - Kim, Mark
AU - Wolf, Matthew
AU - Podhorszki, Norbert
AU - Klasky, Scott
AU - Childs, Hank
AU - Pugmire, David
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2019.
PY - 2019
Y1 - 2019
N2 - This work compares the two major paradigms for doing in situ visualization: in-line, where the simulation and visualization share the same resources, and in-transit, where simulation and visualization are given dedicated resources. Our runs vary many parameters, including simulation cycle time, visualization frequency, and dedicated resources, to study how tradeoffs change over configuration. In particular, we consider simulations as large as 1,024 nodes (16,384 cores) and dedicated visualization resources with as many as 512 nodes (8,192 cores). We draw conclusions about when each paradigm is superior, such as in-line being superior when the simulation cycle time is very fast. Surprisingly, we also find that in-transit can minimize the total resources consumed for some configurations, since it can cause the visualization routines to require fewer overall resources when they run at lower concurrency. For example, one of our scenarios finds that allocating 25% more resources for visualization allows the simulation to run 61% faster than its in-line comparator. Finally, we explore various models for quantifying the cost for each paradigm, and consider transition points when one paradigm is superior to the other. Our contributions inform design decisions for simulation scientists when performing in situ visualization.
AB - This work compares the two major paradigms for doing in situ visualization: in-line, where the simulation and visualization share the same resources, and in-transit, where simulation and visualization are given dedicated resources. Our runs vary many parameters, including simulation cycle time, visualization frequency, and dedicated resources, to study how tradeoffs change over configuration. In particular, we consider simulations as large as 1,024 nodes (16,384 cores) and dedicated visualization resources with as many as 512 nodes (8,192 cores). We draw conclusions about when each paradigm is superior, such as in-line being superior when the simulation cycle time is very fast. Surprisingly, we also find that in-transit can minimize the total resources consumed for some configurations, since it can cause the visualization routines to require fewer overall resources when they run at lower concurrency. For example, one of our scenarios finds that allocating 25% more resources for visualization allows the simulation to run 61% faster than its in-line comparator. Finally, we explore various models for quantifying the cost for each paradigm, and consider transition points when one paradigm is superior to the other. Our contributions inform design decisions for simulation scientists when performing in situ visualization.
UR - http://www.scopus.com/inward/record.url?scp=85067465485&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-20656-7_6
DO - 10.1007/978-3-030-20656-7_6
M3 - Conference contribution
AN - SCOPUS:85067465485
SN - 9783030206550
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 99
EP - 117
BT - High Performance Computing - 34th International Conference, ISC High Performance 2019, Proceedings
A2 - Trinitis, Carsten
A2 - Juckeland, Guido
A2 - Weiland, Michèle
A2 - Sadayappan, Ponnuswamy
PB - Springer Verlag
T2 - 34th International Conference on High Performance Computing, ISC High Performance 2019
Y2 - 16 June 2019 through 20 June 2019
ER -