Serverless 3D Data Optimization Pipelines on AWS

Publication Date: March 8, 2021

Abstract

This implementation guide describes an architecture for scalable 3D Data Optimization Pipelines on AWS using containers, and can be used with any containerized conversion software. An AWS Marketplace Container product, the PiXYZ Scenario Processor, is used as a specific containerized conversion software example. This product enables customers to decimate and simplify 3D Computer Aided Design (CAD) models, and convert them to different industry-standard file formats.

The architecture is designed for horizontal scalability using AWS serverless technologies. Fully-managed AWS services automatically expand and shrink resources based on the customer’s workload demands, so customers don’t need to manage nor provision compute capacity or servers. This allows customers to control costs by only paying for the resources they use, and focus their attention on their 3D Data Optimization requirements instead of infrastructure management. More importantly, customers benefit from automatic scale-out features to handle peak volumes, parallelizing large conversion jobs across the AWS compute cloud. This guide shows how this reduces turn-around times from approximately 14 hours down to approximately 10 minutes.