Computer-aided manufacturing (CAM) software allows for the generation of toolpaths for computer numerical control (CNC) machine tools and enables the creation of sophisticated parts that would not otherwise be possible with conventional manual machining methods. Voxel-based CAM is a recent approach to toolpath planning that enables creation of paths for parts that would be difficult to create with traditional CAM software. However, the use of voxel-based CAM necessitates the presence of powerful hardware (specifically, graphics processing units) in order to perform the necessary computations for creating toolpaths. The concepts of virtualization and desktop-as-a-service offer a promising solution to this challenge, as they allow for many users to access computer hardware that is hosted on a single server. This work investigates the performance impact caused by multiple simultaneous users on voxel-based CAM deployed in a virtualized environment. The implementation of a Python application for multi-user simulation on the virtualized platform is described and timing results gathered from a sequence of simulations are presented and analyzed as the number of users is varied. The results from these simulations demonstrate consistent operational times for a low number of simultaneous users before a period of high performance variation due to resource sharing.
- Manufacturing Engineering Division
Automated Multi-User Analysis of Virtualized Voxel-Based Cam on Shared GPUs
Lynn, R, Medrano, RL, Contis, D, Tucker, T, & Kurfess, T. "Automated Multi-User Analysis of Virtualized Voxel-Based Cam on Shared GPUs." Proceedings of the ASME 2018 13th International Manufacturing Science and Engineering Conference. Volume 2: Materials; Joint MSEC-NAMRC-Manufacturing USA. College Station, Texas, USA. June 18–22, 2018. V002T07A001. ASME. https://doi.org/10.1115/MSEC2018-6322
Download citation file: