-
PatchworkWave: A Multipatch Infrastructure for Multiphysics/Multiscale/Multiframe/Multimethod Simulations at Arbitrary Order
Authors:
Dennis B. Bowen,
Mark Avara,
Vassilios Mewes,
Yosef Zlochower,
Scott C. Noble,
Manuela Campanelli,
Hotaka Shiokawa,
Roseanne M. Cheng,
Julian H. Krolik
Abstract:
We present an extension of the PatchworkMHD code [1], itself an MHD-capable extension of the Patchwork code [2], for which several algorithms presented here were co-developed. Its purpose is to create a multipatch scheme compatible with numerical simulations of arbitrary equations of motion at any discretization order in space and time. In the Patchwork framework, the global simulation is comprise…
▽ More
We present an extension of the PatchworkMHD code [1], itself an MHD-capable extension of the Patchwork code [2], for which several algorithms presented here were co-developed. Its purpose is to create a multipatch scheme compatible with numerical simulations of arbitrary equations of motion at any discretization order in space and time. In the Patchwork framework, the global simulation is comprised of an arbitrary number of moving, local meshes, or patches, which are free to employ their own resolution, coordinate system/topology, physics equations, reference frame, and in our new approach, numerical method. Each local patch exchanges boundary data with a single global patch on which all other patches reside through a client-router-server parallelization model. In generalizing Patchwork to be compatible with arbitrary order time integration, PatchworkMHD and PatchworkWave have significantly improved the interpatch interpolation accuracy by removing an interpolation of interpolated data feedback present in the original Patchwork code. Furthermore, we extend Patchwork to be multimethod by allowing multiple state vectors to be updated simultaneously, with each state vector providing its own interpatch interpolation and transformation procedures. As such, our scheme is compatible with nearly any set of hyperbolic partial differential equations. We demonstrate our changes through the implementation of a scalar wave toy-model that is evolved on arbitrary, time dependent patch configurations at 4th order accuracy.
△ Less
Submitted 31 January, 2020;
originally announced February 2020.
-
Scaling Fine-grained Modularity Clustering for Massive Graphs
Authors:
Hiroaki Shiokawa,
Toshiyuki Amagasa,
Hiroyuki Kitagawa
Abstract:
Modularity clustering is an essential tool to understand complicated graphs. However, existing methods are not applicable to massive graphs due to two serious weaknesses. (1) It is difficult to fully reproduce ground-truth clusters due to the resolution limit problem. (2) They are computationally expensive because all nodes and edges must be computed iteratively. This paper proposes gScarf, which…
▽ More
Modularity clustering is an essential tool to understand complicated graphs. However, existing methods are not applicable to massive graphs due to two serious weaknesses. (1) It is difficult to fully reproduce ground-truth clusters due to the resolution limit problem. (2) They are computationally expensive because all nodes and edges must be computed iteratively. This paper proposes gScarf, which outputs fine-grained clusters within a short running time. To overcome the aforementioned weaknesses, gScarf dynamically prunes unnecessary nodes and edges, ensuring that it captures fine-grained clusters. Experiments show that gScarf outperforms existing methods in terms of running time while finding clusters with high accuracy.
△ Less
Submitted 27 May, 2019;
originally announced May 2019.
-
PATCHWORK: A Multipatch Infrastructure for Multiphysics/Multiscale/Multiframe Fluid Simulations
Authors:
Hotaka Shiokawa,
Roseanne M. Cheng,
Scott C. Noble,
Julian H. Krolik
Abstract:
We present a "multipatch" infrastructure for numerical simulation of fluid problems in which sub-regions require different gridscales, different grid geometries, different physical equations, or different reference frames. Its key element is a sophisticated client-router-server framework for efficiently linking processors supporting different regions ("patches") that must exchange boundary data. T…
▽ More
We present a "multipatch" infrastructure for numerical simulation of fluid problems in which sub-regions require different gridscales, different grid geometries, different physical equations, or different reference frames. Its key element is a sophisticated client-router-server framework for efficiently linking processors supporting different regions ("patches") that must exchange boundary data. This infrastructure may be used with a wide variety of fluid dynamics codes; the only requirement is that their primary dependent variables be the same in all patches, e.g., fluid mass density, internal energy density, and velocity. Its structure can accommodate either Newtonian or relativistic dynamics. The overhead imposed by this system is both problem- and computer cluster architecture-dependent. Compared to a conventional simulation using the same number of cells and processors, the increase in runtime can be anywhere from negligible to a factor of a few; however, one of the infrastructure's advantages is that it can lead to a very large reduction in the total number of zone-updates.
△ Less
Submitted 14 May, 2018; v1 submitted 19 January, 2017;
originally announced January 2017.
