MD hcNewsFlash APU13 – Bigger & Better Than Ever

\

	Put Your Expressions onto a 3D Character With FacePlus in Mixamo All Access, you can capture facial expressions with standard webcams and transfer them in real time onto a 3D character. This real-time capture and animation capability is made possible by offloading Mixamo’s complex algorithms from the CPU onto GPU technology. Real-time motion capture of facial expressions has the potential to make its way into video games and video conferencing applications, giving end users the opportunity to animate avatars in real time.   Did You Miss Hot Chips 2013?  If so, don't despair! Check out these Hot Chips presentations about heterogeneous system architecture (HSA) & HSAIL:   • HSA Queueing - Ian Bratt, ARM   • HSA Memory Model - Ben Gaster, Qualcomm   • HSA Overview - Phil Rogers, AMD   • HSAIL: Portable Compiler for HSA - Ben Sanders, AMD   Developer's Guide for Kabini Processors  If you are a platform designer or involved in the development of low-level BIOS functions, drivers and operating system kernel modules, check out the Preliminary BIOS and Kernel Developer’s Guide (BKDG) for AMD Family 16h Models 00h-0Fh (Kabini) Processors. In this guide, you’ll learn about the major architecture, instruction set and virtualization support changes relative to Family 14h Models 00h-0Fh (ON)   HSA from a Software Developer's Perspective HSA enabled hardware and software improves the programmability of compatible GPUs and accelerators as well as the compute performance for programmers working in mainstream environments. Gary Frost will be exploring these topics on October 14, 2013 at Tsinghua University in Beijing. Gary is an AMD Software Fellow and co-lead of the OpenJDK Sumatra project which promises to bring accelerated GPU compute directly into the JVM. More information about this talk will be available here in the future.   Heterogeneous Computing and HSA in China One Saturday a month, more than 100 developers gather at 3WCoffee in Beijing, China to discuss GPU computing topics such as HSA & GCN, parallelism of image processing and Sparse LU factorization. Check out some of the past presentations from GPU Saturdays here. And starting this month and running through December 2013, presentations about HSA will be given at various universities throughout China. Read this blog post to find the HSA university presentation closest to you.   Asymmetric Multiprocessing with Heterogeneous Architectures: Kurt Shuler of Arteris, Inc. writes in his article titled “Asymmetric Multiprocessing with Heterogeneous Architectures: Use the Best Tool for the Job” about how asymmetric multiprocessing (AMP) architectures make more efficient use of silicon. Read the full article here.

When an AMD Developer Forum user asked about support of templates in OpenCL (Linux), AMD recommended checking out the APP SDK Sample IntrostaticCPP Kernel, which demonstrates how to use templates in OpenCL kernels using AMD's static CPP Kernel extension that is supported in Linux. Visit the AMD Developer Forumsfor other useful tips and tricks. Submit your Tips&Tricks tohcNewsFlash@amd.com AMD is Hiring! Featured Jobs: Architecture Model Designer MTS Software Development Engineer Manager Design Engineering Submit Heterogeneous Compute job listings to hcNewsFlash@amd.com