BlackHoles@Home: BHaHAHA and the Second Binary Black Hole Simulation Campaign

When a gravitational wave (GW) is observed, it is compared against millions of theoretical predictions to estimate parameters and extract important scientific insights. Our most reliable predictions are based on catalogs of numerical relativity (NR) compact binary coalescence simulations, which, until now, have required a computing cluster. This significant computational expense has limited the catalog size to only about 6,000 entries over 15 years. Given the vast parameter space of even the simplest—and most commonly observed—GW source, binary black holes (BBHs), this limited catalog of GW predictions constrains the potential scientific gains from future GW observations.

BlackHoles@Home (BHaH) is a proposed BOINC project that leverages new NR techniques to run BBH simulations on consumer-grade desktop computers, enabling GW follow-ups and catalog expansion with unprecedented throughput using volunteer computers. We recently demonstrated that BHaH's new multi-patch, multi-coordinate numerical gridding algorithms enable it to model BBH inspiral, merger, and ringdown on consumer-grade desktop computers with state-of-the-art accuracy, while requiring only about 1/100th of the memory used by the most popular NR gridding approach, adaptive mesh refinement. We present a two-part BHaH update. First, we introduce BHaHAHA, the BHaH Apparent Horizon Algorithm. BHaHAHA is a simple, robust, and infrastructure-agnostic open-source apparent horizon finder library, built upon the NRPyElliptic hyperbolic relaxation solver. It employs an innovative multigrid-inspired coarse-to-fine resolution approach that accelerates relaxation by more than an order of magnitude, making its performance comparable to AHFinderDirect in the Einstein Toolkit. We also present results from the second BHaH BBH simulation campaign, which includes approximately 100 BBH simulations featuring a significantly upgraded BHaH multipatch algorithm.