Imperial College London
Professor Wayne Luk

ProfessorWayneLuk

Faculty of Engineering, Department of Computing

Professor of Computer Engineering
 
 
 
//

Contact

 

+44 (0)20 7594 8313w.luk Website

 
 
//

Location

 

434Huxley BuildingSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@article{Russell:2017:10.1016/j.cpc.2017.08.011,
author = {Russell, FP and Düben, PD and Niu, X and Luk, W and Palmer, TN},
doi = {10.1016/j.cpc.2017.08.011},
journal = {Computer Physics Communications},
pages = {160--173},
title = {Exploiting the chaotic behaviour of atmospheric models with reconfigurable architectures},
url = {http://dx.doi.org/10.1016/j.cpc.2017.08.011},
volume = {221},
year = {2017}
}
Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - Reconfigurable architectures are becoming mainstream: Amazon, Microsoft and IBM are supporting such architectures in their data centres. The computationally intensive nature of atmospheric modelling is an attractive target for hardware acceleration using reconfigurable computing. Performance of hardware designs can be improved through the use of reduced-precision arithmetic, but maintaining appropriate accuracy is essential. We explore reduced-precision optimisation for simulating chaotic systems, targeting atmospheric modelling, in which even minor changes in arithmetic behaviour will cause simulations to diverge quickly. The possibility of equally valid simulations having differing outcomes means that standard techniques for comparing numerical accuracy are inappropriate. We use the Hellinger distance to compare statistical behaviour between reduced-precision CPU implementations to guide reconfigurable designs of a chaotic system, then analyse accuracy, performance and power efficiency of the resulting implementations. Our results show that with only a limited loss in accuracy corresponding to less than 10% uncertainty in input parameters, the throughput and energy efficiency of a single-precision chaotic system implemented on a Xilinx Virtex-6 SX475T Field Programmable Gate Array (FPGA) can be more than doubled.
AU  - Russell,FP
AU  - Düben,PD
AU  - Niu,X
AU  - Luk,W
AU  - Palmer,TN
DO  - 10.1016/j.cpc.2017.08.011
EP  - 173
PY  - 2017///
SN  - 0010-4655
SP  - 160
TI  - Exploiting the chaotic behaviour of atmospheric models with reconfigurable architectures
T2  - Computer Physics Communications
UR  - http://dx.doi.org/10.1016/j.cpc.2017.08.011
UR  - http://hdl.handle.net/10044/1/50408
VL  - 221
ER  -
Download