High-Precision Numerical Simulations of Rotating Black Holes Accelerated by CUDA
read the original abstract
Hardware accelerators (such as Nvidia's CUDA GPUs) have tremendous promise for computational science, because they can deliver large gains in performance at relatively low cost. In this work, we focus on the use of Nvidia's Tesla GPU for high-precision (double, quadruple and octal precision) numerical simulations in the area of black hole physics -- more specifically, solving a partial-differential-equation using finite-differencing. We describe our approach in detail and present the final performance results as compared with a single-core desktop processor and also the Cell BE. We obtain mixed results -- order-of-magnitude gains in overall performance in some cases and negligible gains in others.
This paper has not been read by Pith yet.
Forward citations
Cited by 3 Pith papers
-
On Computational CUDA Studies of Black Hole Shadows
CUDA-based ray tracing shows black hole shadows and emission rates vary with global monopole, charge, and rotation parameters but are insensitive to the Euler-Heisenberg nonlinearity, yielding observational bounds on ...
-
GPU Acceleration of Collinear and Noncollinear DFT Using a Numerical Atomic Orbital-Based DFT Code
GPU acceleration in OpenMX yields 2.02x speedup for 512-atom collinear DFT and 2.60x for 384-atom noncollinear DFT on two H100 GPUs versus 96 CPU cores.
-
Constraining Black Hole Parameters in Non-Commutative Geometry using Machine Learning
Machine learning constrains non-commutative black hole parameters and reports consistency with Sgr A* Keck observations.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.