6:15-7:00pm: What’s new and first-time attendee intros
7:00-7:45pm: H.265: High Efficiency Video Encoding (Tom Vaughan, MulticoreWare, Inc)
7:55-8:30 pm: Building Your Own GPU Compilers with libNVVM
(Dr Yuan Lin, Senior Engineer, NVIDIA)
Refreshments sponsored by Acceleware!
H.265: High Efficiency Video Encoding
High Efficiency Video Coding (HEVC) is a recently ratified standard for video compression. While HEVC (H.265) can encode video at 30% to 50% lower bit rates than AVC (H.264) at equivalent quality levels, this greater encoding efficiency comes at the cost of greater computational complexity. MulticoreWare is developing a high performance implementation of HEVC, leveraging parallel computing and GPU computing wherever possible.
Speaker: Tom Vaughan
Tom is the VP of Business Development for MulticoreWare, a leader in heterogeneous computing tools, libraries and services. Before joining MulticoreWare, Tom was VP of Products at WinZip, where he was responsible for product strategy and marketing for one of the most popular software applications of all time, with more than 100 million active users. Prior to WinZip, Tom was Director of Business Development at Cyberlink, responsible for US marketing, strategic relationships, and new business development.
Tom started his career as a Mastering Engineer with CBS Records, where he managed mastering operations in one of the first CD manufacturing facilities in the US. Tom holds a B.S. in Computer Engineering and an M.B.A. from Drexel University.
Building Your Own GPU Compilers with libNVVM
The upcoming CUDA 5.5 release will include libNVVM, an LLVM-based solution that makes targeting GPUs from high level languages easier. This talk gives an overview of libNVVM, NVVM IR, libdevice and other related components. We will also walk through several examples showing how to translate generic LLVM IR code to target NVIDIA GPUs. Compiler and/or language developers targeting NVIDIA GPUs and are new to libNVVM will find this presentation useful, so will anyone who just wants to take a peek of what’s under nvcc.
Speaker: Dr Yuan Lin
Yuan Lin is a senior engineer and manages the code generation team at NVIDIA. His team's responsibilities include PTX code generation, tool and platform support. Yuan has been at NVIDIA for 4 years. He was at Sun Microsystems and Motorola before that. He holds a doctorate in computer science from University of Illinois at Urbana-Champaign.
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