| |
Benefits
Overview
Features
Parallel performance
Supported systems
Benefits
- Reduced time-to-market owing to a highly-optimized simulation engine engineered for
high-throughput design assessment on leading-edge computational systems
- Decreased product development costs via highly-accurate algorithms that
allow for virtual prototyping to replace costly physical prototype building and testing
- Increased productivity via design tools engineered with ease-of-use
in mind to facilitate fast learning and rapid deployment
Overview
- 2D and 3D simulation capabilities
- Nonuniform mesh and automesh algorithms
- Simulation convergence autoshutoff
- Parallel / cluster computation
- Movie (.mpg) generation of simulation dynamics
- Scripting language to customize simulation and analysis
- Lorentz, Drude, Debye and anisotropic materials
- Data import/export with BRO's ASAP ray-tracing package
- Data export to Matlab or ASCII file formats
- Structure import from GDSII files
- Extensive online help
Features
- Boundary conditions:
- absorbing (PML), periodic, Bloch, symmetric, asymmetric, and metal boundaries
- Simulation objects:
- primitives that can be rotated and placed in three dimensions, and
structure definition from imported SEM/image files; primitives include
triangles, rectangular blocks, cylinders, conic surfaces, polyons, rings, user-defined
(parametric) surfaces, spheres and pyramids
- Radiation sources:
- waveguide sources, dipoles, plane waves, focused beams and diffraction-limited
spots, total-field scattered-field (TFSF) sources, and source import/export from/to BRO's ASAP
ray-tracing program
- Measurement monitors:
- refractive index monitors, time- and frequency-domain monitors to
measure pulsed or continuous-wave (CW) field profiles and power flow, and movie monitors
to generate .mpg movies of field dynamics
Parallel/clustered performance
The parallel option of FDTD Solutions allows for large-scale and rapid simulation of optical
components, by distributing computational load and memory requirements across multiple
nodes.
Measured data shows that parallel FDTD Solutions offers significant performance enhancements
as measured by the speed increase of simulations performed on multi-core and multiprocessor
computing systems.
Access to an 8-computer cluster will allow you to simulate a structure 8 times larger,
or allow you to simulate the original structure approximately 6 times quicker.
The data further shows that these advantages - the ability to run much larger simulations,
or run simulation much more
quickly - are also available on much larger computer clusters, like the tests run on 128 processors as
shown.
|
|
1
|
The observed speedup is representative of the performance that can be
achieved on systems running multiple jobs simultaneously as with the test system used.
|
|
2
|
All of the results shown as generated using a cluster comprised of
an Intel Xeon 3.0GHz dual-processor machines with 1GB of memory allocated to each processor.
|
Supported systems
FDTD Solutions supports the following systems:
- Vista (including x64 editions), Windows XP, and XP Professsional x64 (64 bit) Edition
- RedHat Enterprise 3, 4 & 5: x86_64 (64 bit) and i386
- SUSE Linux Enterprise 9 & 10: x86_64 (64 bit) and i386
- HPC network technologies including: 100Mb and Gb Ethernet; Myrinet; Infiniband; Infinipath; Quadrics; MPICH and commercial MPI distributions; popular scheduling technologies including PBS Pro, OpenPBS, Condor and Sun Grid Engine.
- Contact Lumerical regarding other supported technologies
FDTD Solutions has limited support on other operating systems that are known to be compatible with the
supported systems above. This includes Windows Server editions, as well as community Linux editions such as
Fedora and OpenSUSE. For operating systems not explicitly listed under supported systems, Lumerical encourages
users to evaluate FDTD Solutions on the desired platform and to contact us for installation support if necessary.
|
|
Home
FDTD Solutions
