Perspectives on High-Performance Computing in a Big Data World: Part D - Learning Model Details and Agent-Based Simulations
17 Oct 2019 | | Contributor(s):: Fox, Geoffrey C.
This lecture completes the discussion of MLforHPC. It covers Learning Model Details and Agents and Time-Series Case Studies.
Perspectives on High-Performance Computing in a Big Data World: Part E - Challenges and Opportunities, Conclusions
This lecture covers the computer science issues raised in this talk. The conclusions note that HPDC/HPC is essential; it is good to work closely with industry with student Internships and Collaborations; the Global AI and Modeling Supercomputer GAIMSC is a good framework with an HPC Cloud linked...
Perspectives on High-Performance Computing in a Big Data World: Part C - MLaroundHPDC/HPC and MLAutotuning
10 Oct 2019 | | Contributor(s):: Fox, Geoffrey C.
This is the first part of the discussion of MLforHPC. It includes MLAutotuning (Using ML to configure or autotune ML or HPC simulations and MLaroundHPC (Learning outputs from inputs).
Perspectives on High-Performance Computing in a Big Data World
30 Sep 2019 | | Contributor(s):: Fox, Geoffrey C.
This course was deleivered at ACM International Symposium on High-Performance Parallel and Distributed Computing (HPDC).High-Performance Computing (HPC) and Cyberinfrastructure have played a leadership role in computational science even since the start of the NSF computing centers program. Thirty...
Perspectives on High-Performance Computing in a Big Data World: Part B - More on the Evolution of Interests and Communities
This part contains several topics. It discusses the importance of industry in several facets of the field: SysML conference, clouds, MLPerf, the Global AI Supercomputer. The nature of data science and data engineering jobs. We emphasize the need for HPC. We finish by introducing MLforHPC (AI for...
Perspectives on High-Performance Computing in a Big Data World: Part A - Data on the Evolution of Interests and Communities
13 Aug 2019 | | Contributor(s):: Fox, Geoffrey C.
This lecture has an overall outline of the 5 part presentation. It covers trends seen from conferences and journals -- the number of papers, attendees and h5index. Then we look at relevant Google Trends. Cyberinfrastructure related activities are less buoyant than those for AI and ML.
How do we solve big science problems using all the modern tools and technologies at our fingertips?
06 Jul 2019 | | Contributor(s):: Jeffrey A. Nichols
Exascale is on the horizon and ORNL just announced our next new system called Frontier to be delivered in 2021 – another order of magnitude more powerful! I will discuss the technologies used in Summit and Frontier.
Darren K Adams
Non-Blocking Conjugate Gradient Methods for Extreme Scale Computing
07 Feb 2016 | | Contributor(s):: Paul Eller
Many scientific and engineering applications use Krylov subspace methods to solve large systems of linear equations. For extreme scale parallel computing systems, the dot products in these methods (implemented using allreduce operations in MPI) can limit performance because they are a...
Range Decomposition: A Low Communication Algorithm for Solving PDEs on Massively Parallel Machines
07 Feb 2016 | | Contributor(s):: Tom Manteuffel
The Range Decomposition (RD) algorithm uses nested iteration and adaptive mesh refinement locally before performing a global communication step. Only several such steps are observed to be necessary before reaching a solution within a small multiple of discretization error. The target application...
A Massively Parallel Semicoarsening Multigrid for 3D Reservoir Simulation on Multi-core and Multi-GPU Architectures
04 Feb 2016 | | Contributor(s):: Abdulrahman Manea
In this work, we have designed and implemented a massively parallel version of the Semicoarsening Black Box Multigrid Solver , which is capable of handling highly heterogeneous and anisotropic 3D reservoirs, on a parallel architecture with multiple GPU’s. For comparison purposes, the...
A Performance Comparison of Algebraic Multigrid Preconditioners on GPUs and MIC
04 Feb 2016 | | Contributor(s):: Karl Rupp
Algebraic multigrid (AMG) preconditioners for accelerators such as graphics processing units (GPUs) and Intel's many-integrated core (MIC) architecture typically require a careful, problem-dependent trade-off between efficient hardware use, robustness, and convergence rate in order to...
HPGMG: Benchmarking Computers Using Multigrid
04 Feb 2016 | | Contributor(s):: Jed Brown
HPGMG (https://hpgmg.org) is a geometric multigrid benchmark designed to measure the performance and versatility of computers. For a benchmark to be representative of applications, good performance on the benchmark should be sufficient to ensure good performance on most important applications and...
Understanding the Propagation of Silent Data Corruption in Algebraic Multigrid
04 Feb 2016 | | Contributor(s):: Jon Calhoun
Sparse linear solvers from a fundamental kernel in high performance computing (HPC). Exascale systems are expected to be more complex than systems of today being composed of thousands of heterogeneous processing elements that operate at near-threshold-voltage to meet power constraints. The...
Preparing for the Future of Computing: The DOE/ASCR Materials Co-Design Center
08 Apr 2015 | | Contributor(s):: Jim Belak
The advent of Advanced / Additive Manufacturing and the Materials Genome Initiative has placed significant emphasis on accelerating the qualification of new materials for use in real applications. Within these workflows lies both the engineering scale qualification through building and testing...
Brenden William Hamilton