File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1109/FCCM.2011.57
- Scopus: eid_2-s2.0-79958734867
- WOS: WOS:000298664800003
Supplementary
- Citations:
- Appears in Collections:
Conference Paper: Mixed precision comparison in reconfigurable systems
Title | Mixed precision comparison in reconfigurable systems |
---|---|
Authors | |
Keywords | reconfigurable system mixed precision |
Issue Date | 2011 |
Citation | Proceedings - IEEE International Symposium on Field-Programmable Custom Computing Machines, FCCM 2011, 2011, p. 17-24 How to Cite? |
Abstract | Customisable data formats provide an opportunity for exploring trade-offs in accuracy and performance of reconfigurable systems. This paper introduces a novel methodology for mixed-precision comparison, which improves comparison performance by using reduced-precision data paths while maintaining accuracy by using high-precision data paths. Our methodology adopts reduced-precision data-paths for preliminary comparison, and high-precision data-paths when the accuracy for preliminary comparison is insufficient. We develop an analytical model for performance estimation of the proposed mixed-precision methodology. Optimisation based on integer linear programming is employed for determining the optimal precision and resource allocation for each of the data paths. The effectiveness of our approach is evaluated using a common collision detection problem. Performance gains of 4 to 7.3 times are obtained over baseline fixed-precision designs for the same FPGAs. With the help of the proposed mixed-precision methodology, our FPGA designs are 15.4 to 16.7 times faster than software running on multi-core CPUs with the same technology. © 2011 IEEE. |
Persistent Identifier | http://hdl.handle.net/10722/200019 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Chow, Gary Chun Tak | - |
dc.contributor.author | Kwok, Kawai | - |
dc.contributor.author | Luk, Wayne | - |
dc.contributor.author | Leong, Philip | - |
dc.date.accessioned | 2014-07-26T23:11:02Z | - |
dc.date.available | 2014-07-26T23:11:02Z | - |
dc.date.issued | 2011 | - |
dc.identifier.citation | Proceedings - IEEE International Symposium on Field-Programmable Custom Computing Machines, FCCM 2011, 2011, p. 17-24 | - |
dc.identifier.uri | http://hdl.handle.net/10722/200019 | - |
dc.description.abstract | Customisable data formats provide an opportunity for exploring trade-offs in accuracy and performance of reconfigurable systems. This paper introduces a novel methodology for mixed-precision comparison, which improves comparison performance by using reduced-precision data paths while maintaining accuracy by using high-precision data paths. Our methodology adopts reduced-precision data-paths for preliminary comparison, and high-precision data-paths when the accuracy for preliminary comparison is insufficient. We develop an analytical model for performance estimation of the proposed mixed-precision methodology. Optimisation based on integer linear programming is employed for determining the optimal precision and resource allocation for each of the data paths. The effectiveness of our approach is evaluated using a common collision detection problem. Performance gains of 4 to 7.3 times are obtained over baseline fixed-precision designs for the same FPGAs. With the help of the proposed mixed-precision methodology, our FPGA designs are 15.4 to 16.7 times faster than software running on multi-core CPUs with the same technology. © 2011 IEEE. | - |
dc.language | eng | - |
dc.relation.ispartof | Proceedings - IEEE International Symposium on Field-Programmable Custom Computing Machines, FCCM 2011 | - |
dc.subject | reconfigurable system | - |
dc.subject | mixed precision | - |
dc.title | Mixed precision comparison in reconfigurable systems | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1109/FCCM.2011.57 | - |
dc.identifier.scopus | eid_2-s2.0-79958734867 | - |
dc.identifier.spage | 17 | - |
dc.identifier.epage | 24 | - |
dc.identifier.isi | WOS:000298664800003 | - |