2014MNRAS.444..961G

The BICEP2 results, when interpreted as a gravitational wave signal and combined with other cosmic microwave background data, suggest a roll-off in power towards small scales in the primordial matter power spectrum. Among the simplest possibilities is a running of the spectral index. Here we show that the preferred level of running alleviates small-scale issues within the ΛCDM model, more so even than viable WDM models. We use cosmological zoom-in simulations of a Milky Way-sized halo along with full-box simulations to compare predictions among four separate cosmologies: a BICEP2-inspired running index model (αs = -0.024), two fixed-tilt ΛCDM models motivated by Planck, and a 2.6 keV thermal WDM model. We find that the running BICEP2 model reduces the central densities of large dwarf-sized haloes (Vmax ∼ 30-80 km/s) and alleviates the too-big-to-fail problem significantly compared to our adopted Planck and WDM cases. Further, the BICEP2 model suppresses the count of small subhaloes by ∼ 50 per cent relative to Planck models, and yields a significantly lower `boost' factor for dark matter annihilation signals. Our findings highlight the need to understand the shape of the primordial power spectrum in order to correctly interpret small-scale data.