Abstract
This paper demonstrates large full-wave inverse-scattering solutions on the Blue Waters supercomputing facility. The high computational complexity of the distorted-Born iterative method (DBIM) is reduced with the multilevel fast multipole algorithm (MLFMA). A hierarchical parallelization of DBIM and MLFMA provides scaling to thousands of computing nodes of the supercomputer. Furthermore, we obtain considerable speedup with modern CPU+GPU architectures over conventional CPU+CPU ones. Results obtained on 4,096 GPU nodes demonstrate a near-real-time tomographic reconstruction, which is crucial for real-life imaging applications.
| Original language | English (US) |
|---|---|
| Title of host publication | 2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 - Proceedings |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 1841-1842 |
| Number of pages | 2 |
| ISBN (Electronic) | 9781538671023 |
| DOIs | |
| State | Published - Jan 1 2018 |
| Event | 2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 - Boston, United States Duration: Jul 8 2018 → Jul 13 2018 |
Publication series
| Name | 2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 - Proceedings |
|---|
Conference
| Conference | 2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 |
|---|---|
| Country | United States |
| City | Boston |
| Period | 7/8/18 → 7/13/18 |
Fingerprint
ASJC Scopus subject areas
- Computer Networks and Communications
- Instrumentation
- Radiation
Cite this
Supercomputing for Full-Wave Tomographic Image Reconstruction in Near-Real Time. / Hidayetoǧlu, Mert; Hwu, Wen-Mei W; Cho Chew, Weng.
2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1841-1842 8608869 (2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 - Proceedings).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Supercomputing for Full-Wave Tomographic Image Reconstruction in Near-Real Time
AU - Hidayetoǧlu, Mert
AU - Hwu, Wen-Mei W
AU - Cho Chew, Weng
PY - 2018/1/1
Y1 - 2018/1/1
N2 - This paper demonstrates large full-wave inverse-scattering solutions on the Blue Waters supercomputing facility. The high computational complexity of the distorted-Born iterative method (DBIM) is reduced with the multilevel fast multipole algorithm (MLFMA). A hierarchical parallelization of DBIM and MLFMA provides scaling to thousands of computing nodes of the supercomputer. Furthermore, we obtain considerable speedup with modern CPU+GPU architectures over conventional CPU+CPU ones. Results obtained on 4,096 GPU nodes demonstrate a near-real-time tomographic reconstruction, which is crucial for real-life imaging applications.
AB - This paper demonstrates large full-wave inverse-scattering solutions on the Blue Waters supercomputing facility. The high computational complexity of the distorted-Born iterative method (DBIM) is reduced with the multilevel fast multipole algorithm (MLFMA). A hierarchical parallelization of DBIM and MLFMA provides scaling to thousands of computing nodes of the supercomputer. Furthermore, we obtain considerable speedup with modern CPU+GPU architectures over conventional CPU+CPU ones. Results obtained on 4,096 GPU nodes demonstrate a near-real-time tomographic reconstruction, which is crucial for real-life imaging applications.
UR - http://www.scopus.com/inward/record.url?scp=85061895210&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85061895210&partnerID=8YFLogxK
U2 - 10.1109/APUSNCURSINRSM.2018.8608869
DO - 10.1109/APUSNCURSINRSM.2018.8608869
M3 - Conference contribution
AN - SCOPUS:85061895210
T3 - 2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 - Proceedings
SP - 1841
EP - 1842
BT - 2018 IEEE Antennas and Propagation Society International Symposium and USNC/URSI National Radio Science Meeting, APSURSI 2018 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
ER -