Small-scale effects of thermal inflation on halo abundance at high-z, galaxy substructure abundance, and 21-cm power spectrum
PHYSICAL REVIEW D
Authors: Hong, Sungwook E.; Zoe, Heeseung; Ahn, Kyungjin
Abstract
We study the impact of thermal inflation on the formation of cosmological structures and present astrophysical observables which can be used to constrain and possibly probe the thermal inflation scenario. These are dark matter halo abundance at high redshifts, satellite galaxy abundance in the Milky Way, and fluctuation in the 21-cm radiation background before the epoch of reionization. The thermal inflation scenario leaves a characteristic signature on the matter power spectrum by boosting the amplitude at a specific wave number determined by the number of e-foldings during thermal inflation (Nbc), and strongly suppressing the amplitude for modes at smaller scales. For a reasonable range of parameter space, one of the consequences is the suppression of minihalo formation at high redshifts and that of satellite galaxies in the Milky Way. While this effect is substantial, it is degenerate with other cosmological or astrophysical effects. The power spectrum of the 21-cm background probes this impact more directly, and its observation may be the best way to constrain the thermal inflation scenario due to the characteristic signature in the power spectrum. The Square Kilometre Array (SKA) in phase 1 (SKA1) has sensitivity large enough to achieve this goal for models with N-bc greater than or similar to 26 if a 10000-hr observation is performed. The final phase SKA, with anticipated sensitivity about an order of magnitude higher, seems more promising and will cover a wider parameter space.
Near-Memory Acceleration for Radio Astronomy
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
Authors: Fiorin, Leandro; Jongerius, Rik; Vermij, Erik; van Lunteren, Jan; Hagleitner, Christoph
Abstract
Processing-in-memory and near-memory computing have recently been rediscovered as a way to alleviate the "memory wall problem" of traditional computing architectures. In this paper, we discuss the implementation of a 3D-stacked near-memory accelerator, targeting radio astronomy and scientific applications. After exploring the design space of the architecture by focusing on minimizing the execution power of the processing pipeline of the SKA1-Low central signal processor, we show that our accelerator can achieve an energy efficiency of up to 390 GFLOPS/W, corresponding to an energy consumption one order of magnitude lower than alternative state-of-the-art implementations. When running additional mathematical and streaming-oriented kernels, our accelerator achieves from 6.4x to 20x energy efficiency improvement compared to alternative solutions.