Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Jie Chen Project Name: ScalaGAUSS Division: MCS Project title: Scalable Gaussian Process Data Analysis Associated funding: Scalable Gaussian Process Analysis of Spatio-Temporal Data (PI: Mihai Anitescu) Other Systems: MCS workstations Science: This project uses a maximum likelihood approach to perform Gaussian process analysis on both synthetic and real data. The observations considered for analysis are simulation and data from nuclear engineering and climate science. The resulting analysis is used to describe the spatial correlation of the sampled data, to predict and to understand uncertainty in very high fidelity phenomena. Project description: ScalaGAUSS is a statistical software for analyzing spatialtemporal data arising from various applications (chemistry, power grids, material science) and modeled as a Gaussian process. The goal of ScalaGAUSS is to handle both gridded and scattered data at the scale of 10^9 to 10^12 points and beyond. Following the framework of a reformulation of the maximum likelihood estimation using sample average approximation, the core of the computations is to solve several linear systems with respect to the fully dense covariance matrix (having the number of rows/columns the same the data size). For gridded data, we implemented a multilevel Toeplitz linear solver and tested it with up to 1024 cores last year; see the FY2012 report. This year’s work will be built on that of the last year. In particular, we will be focusing on 1. Improving the scaling of the solver; 2. Incorporating the solver into ScalaGAUSS. The objective is to finish a working version of ScalaGAUSS that can be applied to real applications, such as for analyzing ozone data. On the other hand, for scatted data, a linear solver requires efficient matrix-vector multiplications. For this, we have developed a fast summation program for a specific covariance kernel, based on the treecode methodology. This program will be the essential building block of the linear solver. We will be investigating its performance for solving linear systems. The goal is to add to ScalaGAUSS a component that handles scattered data as efficiently as for gridded data. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 400000 Q1: 100000 Q2: 100000 Q3: 100000 Q4: 100000 Justification: Thank You, The LCRC Accounts System