Techniques for implementation of the FDTD method on a CM-5 parallel computer

Liu, Z.M.; Mohan, A.S.; Aubrey, Tim; Belcher, W.R.

doi:10.1109/74.475869

Cited by 26 publications

(17 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, the model is limited in size due to the simulation time and amount of memory required to simulate larger devices. Methods to ease the requirements of three-dimensional models include use of Graphics Processing Units (GPU), parallel processing, and mode simplifications [21][22][23]. Since threedimensional model parameters scale in simulation time like N4 and in memory like N3, the two-dimensional FDTD is used in this paper to simulate this model to avoid unnecessary complications [24].…”

Section: Simulation Modelmentioning

confidence: 99%

Improvement of solar cell efficiency using nano-scale top and bottom grating

2011

View full text Add to dashboard Cite

We study solar-cell designs using nano-grating on both top (transmission) and bottom (reflection) of the solar cell. First, we perform simulations based on rigorous coupled wave analysis (RCWA) to evaluate the diffraction top gratins. In RCWA method, we calculate up to 20 harmonics, and sweep the launch angle of incident light from 0 to 90 degree. The incident light varies from100nm to 1200nm wavelength. Triangular grating can achieve higher light absorption compared to the rectangular grating. The best top grating is around 200nm grating period, 100nm grating height, and 50% filling factor, which responses to 37% improvement for triangular grating and 23% for rectangular grating compared to non-grating case. Then, we use Finite-Difference Time-Domain (FDTD) to simulate transmission/reflection double grating cases. We simulated triangular-triangular (top-bottom) grating cases and triangular-rectangular (topbottom) grating case. We realize solar cell efficiency improvement about 42.4%. For the triangulartriangular (top-bottom) grating case, the 20% efficiency improvement is achieved. Finally, we present weighted-light simulation for the double grating for the first time and show the best grating can achieve 104% light improvement, which is quite different from traditional non-weighted calculation.

show abstract

Section: Simulation Modelmentioning

confidence: 99%

Improvement of solar cell efficiency using nano-scale top and bottom grating

2011

View full text Add to dashboard Cite

show abstract

“…Liu et al [8] parallelize the core FDTD on a CM-5 (32 processors) parallel computer, with a performance gain of 113 seconds for a domain of 96×86×102 in 1500 time steps over nearly 189 minutes by the sequential version. Guiffaut et al [5] implement a parallel FDTD on a computational domain of 150 × 150 × 50 cells on Cray T3E.…”

Section: Introductionmentioning

confidence: 99%

Finite-difference time-domain on the cell/B.E. processor

Thulasiraman

2008

2008 IEEE International Symposium on Parallel and Distributed Processing

View full text Add to dashboard Cite

Finite-Difference Time-Domain (FDTD) is a kernel usedto solve problems in electromagnetics applications such as microwave tomography. It is a data-intensive and computation-intensive problem. However, its computation scheme indicates that an architecture with SIMD support has the potential to bring performance improvement over traditional architectures without SIMD support. The Cell Broadband Engine (Cell/B.E.) processor is an implementation of a heterogeneous multicore architecture. It consists of one conventional microprocessor, PowerPC Processor Element (PPE), and eight SIMD co-processor elements, Synergistic Processor Elements (SPEs). One unique feature of an SPE is that it has 128-entry 128-bit uniform registers which support SIMD. Therefore, FDTD may be mapped well on Cell/B.E. processor. However, each SPE can directly access only 256KB local store (LS) both for instructions and data. The size of LS is much less than what is needed for an accurate simulation of FDTD which requires large number of fine-grained Yee cells. In this paper, we design the algorithm on Cell/B.E. by efficiently using the asynchronous DMA (direct memory access) mechanism available on an SPE transferring data between its LS and the main memory via the high bandwidth bus on-chip EIB (Element Interconnect Bus). The new algorithm was run on an IBM Blade QS20 blades running at 3.2GHz. For a computation domain of 600 × 600 Yee cells, we achieve an overall speedup of 14.14 over AMD Athlon and 7.05 over AMD Opteron at the processor level.

show abstract

“…FDTD(Finite Difference Time Domain) (1) (2) FDTD Maxwell Green (3) (4) , 2 4 FDTD (5) ADI FDTD (6) HP-735 8 10Base-T WS (8) Cray T3E HITACHI SR8000 (9) (11) (1) …”

mentioning

confidence: 99%

Large Scale EM Wave Propagation Analysis using FDTD Parallel Computation on Computer System for Education

Sonoda

2004

IEEJ Trans.EIS

View full text Add to dashboard Cite

This paper describes the study of a fast electromagnetic (EM) wave propagation analysis that can solve electrically large domains using finite difference time domain (FDTD) method on cluster of personal computers (PC cluster). It reports an implementation of parallel FDTD using an MPI library on PC clusters of the computer system for education. Use of this method demonstrates that the speed-up ratio achieved for problem size 1200 × 1200 is about 55.0 using FDTD on 80 PCs. And also, indoor propagation of UWB pulse on the floor (1095.4λ × 98.6λ) is analyzed by the parallel FDTD using 40 PCs, computational time and memory have been reduced by 1/36.4 and 1/39.9, respectively. The results demonstrate that the parallel FDTD using PC cluster can analyze electrically large problems low computational costs than novel FDTD.

show abstract

Techniques for implementation of the FDTD method on a CM-5 parallel computer

Cited by 26 publications

References 15 publications

Improvement of solar cell efficiency using nano-scale top and bottom grating

Improvement of solar cell efficiency using nano-scale top and bottom grating

Finite-difference time-domain on the cell/B.E. processor

Large Scale EM Wave Propagation Analysis using FDTD Parallel Computation on Computer System for Education

Contact Info

Product

Resources

About