Efficient Software Architecture Pattern for Accelerator Based Computing

Graphics Accelerators are increasingly used for general purpose high performance computing applications as they provide a low cost solution to high performance computing requirements. Intel also came out with a performance accelerator that offers a similar solution. However, the existing application software needs to be restructured to suit to the accelerator paradigm with a suitable software architecture pattern. In the present work, master-slave architecture is employed to convert CFD grid free Euler solvers in CUDA for GPGPU computing. The performance obtained using master-slave architecture for GPGPU computing is compared with that of sequential computing results.

Test Frame Design for the Characterization of Additive Manufacturing Compliant Materials

This work presents a design of a 6 degree of freedom (DOF) robotic test frame designed to provide multiple and combined loading scenarios for additive manufacturing (AM) materials. The need is to provide a more in-depth look into the material properties of nonlinear anisotropic materials as traditional uniaxial or biaxial test frames have been shown to be inefficient in providing accurate material property values. With the application of surrogate models with General Purpose Graphics Processing (GPGPU) computing, “real-time” characterization is achievable. The work provided is a next generation 6 DOF test frame designed to reducing costs, increasing workspace, and reducing overall size over previous designs.