Optical Design On Steroids

This is the second in a four-part series on "design for manufacturing and assembly" or DFMA for short — an innovative way to bring prototypes, or even cocktail-napkin concepts, into profitable production.

By linking computers together in parallel, cluster computing unleashes an enormous simulation capability. As a result, it can optimize your design, manufacturing, and assembly process for optical components – before any mould is cast or any lens is ground. Cluster computing puts traditional optical design on steroids.

In ELCAN'S unique version of cluster computing for optical design, its developers are using MATLAB®, a programming language for high level technical computing, coupled with CODE V® optical design software. Together they can harness the power of more than 100 computers for optical design processes such as tolerancing and optimization. Design and production optimisation

In order to determine the ideal production process, the ELCAN computer cluster runs a "Monte Carlo" simulation of optical design tolerances on the customer's specifications. A subsequent engineering analysis delivers the most favourable optical design for manufacturability and assembly.

"Monte Carlo simulations are traditionally slow – one "single-core" computer can run approximately 200 simulations per hour," explains Wolf Glage, Vice President of Engineering for ELCAN Optical Technologies. "ELCAN's "multi-core" computer cluster can run approximately 10,000 trials in this same timeframe. These results provide higher statistical significance, with results at least an order of magnitude more accurate."

The net result – your cocktail napkin concept gets to market on time and on budget.