Photon Design представила инструмент для моделирования кольцевых резонаторов с точностью FDTD и скоростью в 100 раз выше

Photon Design's MT-FIMMPROP tool can now simulate ring resonators Photon Design’s multi-topology MT-FIMMPROP photonic simulation tool, which uses a patented a EigenMode Expansion (EME) computational method, can now simulate ring resonator designs - a first in the photonics industry. The company quantified the performance of the new EME capability in a recent white paper, in which it concluded that EME delivered equivalent accuracy to industry-standard, Finite Difference Time Domain (FDTD) tools, but was over 100 times faster using a workstation PC, with no cloud processing necessary. Engineers run many simulations to fine-tune photonics designs, so the shorter runtimes provided by EME enable practical, iterative workflows. In addition, the efficiency of the MT-FIMMPROP software itself, also plays a valuable part in engineers achieving optimal, photonic design results. Dominic Gallagher, CEO of Photon Design, said: “Photon Design’s MT-FIMMPROP is the only EME-based simulation tool to successfully model ring resonators. Our EME delivers the accuracy of FDTD at far faster simulation speeds and without the need for expensive, pay-per-simulation, cloud computing; often needed for FDTD. “Our recently published white paper showed that MT-FIMMPROP can process a typical 30um-radius ring resonator in just 11 seconds, on a laptop PC. By comparison, FDTD took four hours for a much smaller and simpler, 12um-radius ring resonator, using a workstation PC. FDTD is slower because it requires more than 11 trillion calculations to process the 12um example, with its computational load increasing eight-fold every time the device size doubles. Had FDTD been used for the 30um ring resonator, the simulation would have taken 15 times longer again. “MT-FIMMPROP is an efficient computational modeller. Unlike FDTD, which processes the entire device bounding box, EME only processes the ring-resonators’ computational regions, which typically make up less than 5% of the device volume. It also enables engineers to combine and reuse repeated computational regions to avoid both design and process duplication. This simplifies the design process, reduces runtimes and actively supports an iterative photonic design approach.”