English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
 
 
DownloadE-Mail
  From Physics Model to Results: An Optimizing Framework for Cross-Architecture Code Generation

Blazewicz, M., Hinder, I., Koppelman, D. M., Brandt, S. R., Ciznicki, M., Kierzynka, M., et al. (2013). From Physics Model to Results: An Optimizing Framework for Cross-Architecture Code Generation. Scientific Programming, 21, 1-16. doi:10.3233/SPR-130360.

Item is

Files

show Files
hide Files
:
1307.6488.pdf (Preprint), 843KB
Name:
1307.6488.pdf
Description:
File downloaded from arXiv at 2013-08-14 12:42
OA-Status:
Visibility:
Public
MIME-Type / Checksum:
application/pdf / [MD5]
Technical Metadata:
Copyright Date:
-
Copyright Info:
-
:
SP21_1.pdf (Any fulltext), 485KB
Name:
SP21_1.pdf
Description:
-
OA-Status:
Visibility:
Public
MIME-Type / Checksum:
application/pdf / [MD5]
Technical Metadata:
Copyright Date:
-
Copyright Info:
-
License:
-

Locators

show

Creators

show
hide
 Creators:
Blazewicz, Marek1, Author
Hinder, Ian2, Author           
Koppelman, David M., Author
Brandt, Steven R., Author
Ciznicki, Milosz, Author
Kierzynka, Michal, Author
Löffler, Frank, Author
Schnetter, Erik, Author
Tao, Jian, Author
Affiliations:
1Poznań Supercomputing and Networking Center Poznań University of Technology, ou_persistent22              
2Astrophysical Relativity, AEI-Golm, MPI for Gravitational Physics, Max Planck Society, ou_24013              

Content

show
hide
Free keywords: Physics, Computational Physics, physics.comp-ph,Computer Science, Mathematical Software, cs.MS,General Relativity and Quantum Cosmology, gr-qc
 Abstract: Starting from a high-level problem description in terms of partial differential equations using abstract tensor notation, the Chemora framework discretizes, optimizes, and generates complete high performance codes for a wide range of compute architectures. Chemora extends the capabilities of Cactus, facilitating the usage of large-scale CPU/GPU systems in an efficient manner for complex applications, without low-level code tuning. Chemora achieves parallelism through MPI and multi-threading, combining OpenMP and CUDA. Optimizations include high-level code transformations, efficient loop traversal strategies, dynamically selected data and instruction cache usage strategies, and JIT compilation of GPU code tailored to the problem characteristics. The discretization is based on higher-order finite differences on multi-block domains. Chemora's capabilities are demonstrated by simulations of black hole collisions. This problem provides an acid test of the framework, as the Einstein equations contain hundreds of variables and thousands of terms.

Details

show
hide
Language(s):
 Dates: 2013-07-242013
 Publication Status: Issued
 Pages: 18 pages, 4 figures, accepted for publication in Scientific Programming
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Degree: -

Event

show

Legal Case

show

Project information

show

Source 1

show
hide
Title: Scientific Programming
Source Genre: Journal
 Creator(s):
Affiliations:
Publ. Info: -
Pages: - Volume / Issue: 21 Sequence Number: - Start / End Page: 1 - 16 Identifier: -