A hardware-accelerated graphics processing unit-based computational machine will be designed, developed, and commissioned to support reduction and analysis of time series of stellar spectra. The goal of this project is to study the variability to infer details of the structure of the inside of the observed stars. This tool, referred to as asteroseismology, is a unique method to probe a star's internal structure from measurements of its discrete oscillation spectrum. Asteroseismology was originally developed for studying the sun, but advances in space-based observations are allowing its application to other stars.
The principal goal of this project was to design, develop, commission and deploy a GPU-accelerated spectral synthesis engine for modeling line profiles as a component of asteroseismic mode identification. A spectral synthesis engine is a computer platform, combining both software and hardware, for predicting the spectrum of electromagnetic radiation emitted by a star. Typically this calculation is extremely time consuming, as it involves huge numbers of interpolations in pre-calculated tabulations of stellar-atmosphere radiant intensities. However, the new engine developed by the project, dubbed GRASSY (GRaphics processing unit-Accelerated Spectral SYnthesis), uses graphics processing units (GPUs) together with custom software to greatly accelerate the spectral synthesis calculations. GPUs, the hardware components in modern PCs which handle image manipulation and display, contain specialized 'texture units' for performing table interpolations; by taking advantage of these texture units, GRASSY will be able to perform spectral synthesis in a fraction of the time it takes ordinary central processing units (CPUs). GRASSY is specifically aimed at synthesizing spectral line profiles of oscillating stars. Comparing these synthetic profiles against time-series observations of actual stars is the key component of 'asteroseismic mode identification' --- determining which specific oscillation modes are excited in a given star. This mode identification is important as it allows us to probe the star's interior structure, in much the same way that earthquakes and tremors are used be terrestrial seismologists to determine the Earth's internal structure. Once deployed, GRASSY will be made publicly accessible through a Guest Analyst program, allowing the engine to be used for mode identification by the wider scientific community. The GRASSY engine was not completed during the timeframe of the project, but development is still ongoing and the engine will be open for public access in 2014.
