A precompiler to reduce the memory footprint of multiscale PDE solvers in C++
| Autor: | Christoph Zenger, Tobias Weinzierl, Hans-Joachim Bungartz, Wolfgang Eckhardt |
|---|---|
| Rok vydání: | 2010 |
| Předmět: |
Multi-core processor
Flat memory model Computer Networks and Communications Computer science Uniform memory access Overlay Parallel computing Extended memory Memory address Memory management Shared memory Hardware and Architecture CUDA Pinned memory Memory footprint Interleaved memory Code generation Distributed memory Software |
| Zdroj: | Future Generation Computer Systems. 26:175-182 |
| ISSN: | 0167-739X |
| DOI: | 10.1016/j.future.2009.05.011 |
| Popis: | A PDE solver's value is increasingly co-determined by its memory footprint, as the increase of computational multicore power overtakes the memory access speed, and as memory restricts the maximum experiment size. Tailoring a code to require less memory is technical challenging, error-prone, and hardware-dependent. Object-oriented code typically consumes much memory, though developers favour such high-level languages offering meaningful models and good maintainability. We augment the language C++ with new keywords branding records to be memory-critical. Our precompiler DaStGen then transforms this augmented specification into plain C++ optimised for low memory requirements. Hereby, it encodes multiple attributes with fixed range within one variable, and it reduces the number of bits per floating point value. The tool also generates one user-defined MPI data type per class and, thus, facilitates the construction of parallel codes with small messages. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect