Come for Table discussions, Member Self-Intro, What's New, Application Showcase, and Advanced Application Development Techniques! Exchange ideas, meet experts, share code... all HPC & GPU, all practical, all cutting-edge.

Agenda:

General Discussions:

6:15-6:30pm    What’s new in HPC & GPU Supercomputing

6:30-6:45pm    Member self-intros: 30 seconds for each member

Main Program:

6:45-7:45pm  Supporting Intelligent Robotics with Parallel Computation combining Coarse, Medium and Fine Grained Parallel Computations (Robert Rossi)

7:45-7:50pm  Break

7:50-8:20pm  Challenges in supporting multiple HW vendors and multiple GPU programming languages while enabling multi-GPU Deployments - A case study on SIMULIA Abaqus (Dan Cyca)

Book Review:

8:20-8:30pm Selected Chapters

Refreshments courtesy of Acceleware

Main Talk: Supporting Intelligent Robotics with Parallel Computation combining Coarse, Medium and Fine Grained Parallel Computations

Abstract: XOTAR is developing a processing platform for intelligent or autonomous robots, with a heavy emphasis on Vision and Perception Action Systems. The evolving platform architecture remotes a major portion of the computation for a robot between the robots embedded sensors/controls and a cloud based GP/GPU system using compressed sensing and bidirectional streaming. It also allows simulation based reasoning to be performed in the cloud, enabling offline learning of optimal environment adaptations. This application focused presentation will discuss the background, challenges and accomplishments achieved in using GP/GPU and multigrain parallel processing to enable future autonomous robots and cluster robots (multi robots that behave as a single system).

Bio: Robert Rossi is founder of XOTAR, a Saratoga CA based startup working on human and animal inspired intelligent robots. Before founding XOTAR in 2009, he led the successful campaign to deliver the successor to the JPEG image coding standard and led the Core Media Processing Team at Microsoft. JEPG’s successor, called JPEG XR, is used in XOTAR’s vision compression stream technology.  Robert is a 30 year veteran who founded his first company at age 21 in Smart Cameras/Computer Vision. His career focus has centered in AI and High Performance Computing applied to banking, defense and consumer electronics. 

Short Talk: Challenges in supporting multiple HW vendors and multiple GPU programming languages while enabling multi-GPU Deployments - A case study on SIMULIA Abaqus

Abstract: SIMULIA Abaqus is a finite-element analysis product suite.  Acceleware worked with SIMULIA to add GPU acceleration capabilities to the direct sparse solver.  The solver supports NVIDIA GPUs via CUDA, and AMD GPUs via OpenCL.  Multiple GPU support is currently in beta.  We will discuss the challenges in supporting multiple HW vendors and multiple GPU programming languages, and in enabling multiple GPU support.

Bio: Dan Cyca, Chief Technical Officer, Acceleware Ltd.

Regarded as a leading mind in the field of parallel processing, Dan has extensive experience working with GPUs, clusters and multi-core solutions. Dan joined Acceleware in 2004 as a software developer to build the company’s first product. Since then, he has served in many technical and leadership roles in the company. Most recently, as the Director of Engineering, Dan was responsible for managing the software development group. Prior to Acceleware, Dan’s experience included developing 'C-to-hardware' compilers, and implementing digital signal processing and encryption algorithms on FPGAs. Dan has an M. Sc. in Electrical Engineering from the University of Calgary.

Location:

Room 109/110; 
Carnegie Mellon Silicon Valley; 
NASA Research Park Bldg 23; 
Mountain View, CA 94043; 

Directions to Carnegie Mellon Silicon Valley; 

Google Map showing parking, check point, and building entrance; 

NOTE: You will need a government issued ID (e.g. Driver's License) to enter NASA Research Park