Xevolver: A code transformation framework for separation of system‐awareness from application codes

2019 ◽  
Vol 32 (7) ◽  
Author(s):  
Kazuhiko Komatsu ◽  
Ayumu Gomi ◽  
Ryusuke Egawa ◽  
Daisuke Takahashi ◽  
Reiji Suda ◽  
...  
Keyword(s):  
Code Transformation ◽  
Application Codes

scholarly journals MiniApps derived from production HPC applications using multiple programing models

2016 ◽  
Vol 32 (4) ◽  
pp. 582-593 ◽  
Author(s):  
OE Bronson Messer ◽  
Ed D’Azevedo ◽  
Judy Hill ◽  
Wayne Joubert ◽  
Mark Berrill ◽  
...  
Keyword(s):  
Large Scale ◽  
Performance Characteristics ◽  
Production Mode ◽  
Actual Performance ◽  
Oak Ridge ◽  
Application Codes

We have developed a set of reduced, proxy applications (“MiniApps”) based on large-scale application codes supported at the Oak Ridge Leadership Computing Facility (OLCF). The MiniApps are designed to encapsulate the details of the most important (i.e. the most time-consuming and/or unique) facets of the applications that run in production mode on the OLCF. In each case, we have produced or plan to produce individual versions of the MiniApps using different specific programing models (e.g., OpenACC, CUDA, OpenMP). We describe some of our initial observations regarding these different implementations along with estimates of how closely the MiniApps track the actual performance characteristics (in particular, the overall scalability) of the large-scale applications from which they are derived.

scholarly journals Continuous Delivery of Blockchain Distributed Applications

Sensors ◽  
2021 ◽  
Vol 22 (1) ◽  
pp. 128
Author(s):  
Tomasz Górski
Keyword(s):  
Unified Modeling Language ◽  
Electricity Consumption ◽  
Distributed Applications ◽  
Code Transformation ◽  
Network Nodes ◽  
Unified Modeling ◽  
Distributed Ledger ◽  
Business Application ◽  
Continuous Delivery ◽  
Execution Environment

Ensuring a production-ready state of the application under development is the imminent feature of the Continuous Delivery (CD) approach. In a blockchain network, nodes communicate and store data in a distributed manner. Each node executes the same business application but operates in a distinct execution environment. The literature lacks research focusing on continuous practices for blockchain and Distributed Ledger Technology (DLT). Specifically, it lacks such works with support for both design and deployment. The author has proposed a solution that takes into account the continuous delivery of a business application to diverse deployment environments in the DLT network. As a result, two continuous delivery pipelines have been implemented using the Jenkins automation server. The first pipeline prepares a business application whereas the second one generates complete node deployment packages. As a result, the framework ensures the deployment package in the actual version of the business application with the node-specific up-to-date version of deployment configuration files. The Smart Contract Design Pattern has been used when building a business application. The modeling aspect of blockchain network installation has required using Unified Modeling Language (UML) and the UML Profile for Distributed Ledger Deployment. The refined model-to-code transformation generates deployment configurations for nodes. Both the business application and deployment configurations are stored in the GitHub repositories. For the sake of verification, tests have been conducted for the electricity consumption and supply management system designed for prosumers of renewable energy.

scholarly journals AGS to RHIC Beam Line: Application Codes

1994 ◽  
Author(s):  
W. MacKay ◽  
T. Satogata
Keyword(s):  
Beam Line ◽  
Application Codes

scholarly journals Exploring the feasibility of lossy compression for PDE simulations

2018 ◽  
Vol 33 (2) ◽  
pp. 397-410 ◽  
Author(s):  
Jon Calhoun ◽  
Franck Cappello ◽  
Luke N Olson ◽  
Marc Snir ◽  
William D Gropp
Keyword(s):  
Energy Cost ◽  
High Performance ◽  
Truncation Error ◽  
Computational Cost ◽  
Lossy Compression ◽  
Data Movement ◽  
Hardware Failure ◽  
Traditional Approaches ◽  
Application Codes ◽  
Performance Computing

Checkpoint restart plays an important role in high-performance computing (HPC) applications, allowing simulation runtime to extend beyond a single job allocation and facilitating recovery from hardware failure. Yet, as machines grow in size and in complexity, traditional approaches to checkpoint restart are becoming prohibitive. Current methods store a subset of the application’s state and exploit the memory hierarchy in the machine. However, as the energy cost of data movement continues to dominate, further reductions in checkpoint size are needed. Lossy compression, which can significantly reduce checkpoint sizes, offers a potential to reduce computational cost in checkpoint restart. This article investigates the use of numerical properties of partial differential equation (PDE) simulations, such as bounds on the truncation error, to evaluate the feasibility of using lossy compression in checkpointing PDE simulations. Restart from a checkpoint with lossy compression is considered for a fail-stop error in two time-dependent HPC application codes: PlasComCM and Nek5000. Results show that error in application variables due to a restart from a lossy compressed checkpoint can be masked by the numerical error in the discretization, leading to increased efficiency in checkpoint restart without influencing overall accuracy in the simulation.

Sign in / Sign up

Export Citation Format

Share Document