Abstract
We report inverse-scattering solutions on supercomputers involving large numbers of graphics processing units (GPUs). The distorted-Born iterative method (DBIM) is employed for the iterative inversions. In each iteration, the required forward problems are distributed among computing nodes equipped with GPUs, and solved with the multilevel fast multipole algorithm. A tomographic reconstruction of a synthetic object with a linear dimension of one hundred wavelengths is obtained on 256 GPUs. The results show that DBIM obtains images approximately four times faster on GPUs, compared to parallel executions on traditional CPU-only computing nodes.
| Original language | English (US) |
|---|---|
| Title of host publication | 2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017 |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| ISBN (Electronic) | 9780996007832 |
| DOIs | |
| State | Published - May 1 2017 |
| Event | 2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017 - Firenze, Italy Duration: Mar 26 2017 → Mar 30 2017 |
Publication series
| Name | 2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017 |
|---|
Other
| Other | 2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017 |
|---|---|
| Country | Italy |
| City | Firenze |
| Period | 3/26/17 → 3/30/17 |
Fingerprint
ASJC Scopus subject areas
- Radiation
- Signal Processing
- Computational Mathematics
- Instrumentation
Cite this
Large inverse-scattering solutions with DBIM on GPU-enabled supercomputers. / Hidayetogglu, Mert; Pearson, Carl; Chew, Weng Cho; Gurel, Levent; Hwu, Wen-Mei W.
2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017. Institute of Electrical and Electronics Engineers Inc., 2017. 7916310 (2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Large inverse-scattering solutions with DBIM on GPU-enabled supercomputers
AU - Hidayetogglu, Mert
AU - Pearson, Carl
AU - Chew, Weng Cho
AU - Gurel, Levent
AU - Hwu, Wen-Mei W
PY - 2017/5/1
Y1 - 2017/5/1
N2 - We report inverse-scattering solutions on supercomputers involving large numbers of graphics processing units (GPUs). The distorted-Born iterative method (DBIM) is employed for the iterative inversions. In each iteration, the required forward problems are distributed among computing nodes equipped with GPUs, and solved with the multilevel fast multipole algorithm. A tomographic reconstruction of a synthetic object with a linear dimension of one hundred wavelengths is obtained on 256 GPUs. The results show that DBIM obtains images approximately four times faster on GPUs, compared to parallel executions on traditional CPU-only computing nodes.
AB - We report inverse-scattering solutions on supercomputers involving large numbers of graphics processing units (GPUs). The distorted-Born iterative method (DBIM) is employed for the iterative inversions. In each iteration, the required forward problems are distributed among computing nodes equipped with GPUs, and solved with the multilevel fast multipole algorithm. A tomographic reconstruction of a synthetic object with a linear dimension of one hundred wavelengths is obtained on 256 GPUs. The results show that DBIM obtains images approximately four times faster on GPUs, compared to parallel executions on traditional CPU-only computing nodes.
UR - http://www.scopus.com/inward/record.url?scp=85020032006&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85020032006&partnerID=8YFLogxK
U2 - 10.23919/ROPACES.2017.7916310
DO - 10.23919/ROPACES.2017.7916310
M3 - Conference contribution
T3 - 2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017
BT - 2017 International Applied Computational Electromagnetics Society Symposium - Italy, ACES 2017
PB - Institute of Electrical and Electronics Engineers Inc.
ER -