Presentation | 2011-04-12 A Note on Data Compression of Double-Precision Floating-Point Numbers for Massively Parallel Numerical Simulations Mamoru OHARA, Takashi YAMAGUCHI, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In numerical simulations using massively parallel computers like GPGPU (General-Purpose computing on Graphics Processing Units), we often need to transfer computational results from external devices such as GPUs to the main memory or secondary storage of the host machine in order to display them or save them for check- pointing. Since size of the computation results is sometimes unacceptably large to hold them, it is desired that the data is compressed and stored; this is especially true for checkpoint data, which is used only when failures occur. In addition, considering overheads for transferring data between the devices and host memories, it is preferable that the data is compressed in a part of parallel computation performed on the devices. Traditional compression methods for floating-point numbers do not always show good parallelism. In this paper, we propose a new compression method for massively-parallel simulations, in which we predict floating-point numbers with polynomial extrapolation. We also present numerical examples of compression ratio and time obtained from experimental implementations of the proposed method runnig on GPUs. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | massively-parallel numerical simulations / GPGPU / checkpointing / double-precision floating-point numbers / lossless compression |
Paper # | CPSY2011-2,DC2011-2 |
Date of Issue |
Conference Information | |
Committee | DC |
---|---|
Conference Date | 2011/4/5(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Dependable Computing (DC) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | A Note on Data Compression of Double-Precision Floating-Point Numbers for Massively Parallel Numerical Simulations |
Sub Title (in English) | |
Keyword(1) | massively-parallel numerical simulations |
Keyword(2) | GPGPU |
Keyword(3) | checkpointing |
Keyword(4) | double-precision floating-point numbers |
Keyword(5) | lossless compression |
1st Author's Name | Mamoru OHARA |
1st Author's Affiliation | Tokyo Metropolitan Industrial Technology Research Institute() |
2nd Author's Name | Takashi YAMAGUCHI |
2nd Author's Affiliation | Tokyo Metropolitan Industrial Technology Research Institute |
Date | 2011-04-12 |
Paper # | CPSY2011-2,DC2011-2 |
Volume (vol) | vol.111 |
Number (no) | 2 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |