Acceleware Corp announced their fall 2010 comprehensive CUDA/OpenCL training schedule co-sponsored by Microsoft Corp. Acceleware's training classes are designed to support the HPC communities GPU programmers using CUDA and OpenCL, along with instruction on Microsoft's HPC Server 2008 cluster operating systems. There are currently five course schedule between mid-October and mid-December, with each course lasting 5 days. You can find an overview of the courses on our Khronos Group events page. Complete details are available on the Acceleware training website.
AMD Developer Central has a new section that offers OpenCL Code Samples. Although only one sample right now, this looks like a promising page for those getting started with OpenCL. The Khronos Group also has a section with numerous OpenCL samples and tutorials on their site.
OpenCL framework to accelerate an EMRI modeling application using the hardware accelerators – Cell BE and Tesla CUDA GPU. The main goal of this work is to evaluate an emerging computational platform, OpenCL, for scientific computation. Results show OpenCL binary on a par with CUDA SDK. Baseline is an AMD Phenom 2.5Ghz CPU.
In this episode of Adventures in OpenCL tutorials, we cover OpenCL context sharing with OpenGL. We make a simple particle system to demonstrate this feature. One of the most important aspects of this feature is the time we can save by doing rendering and calculations on the same memory in the GPU, this means we don’t need to copy data back and forth!
AMD Developer central just released an OpenCL optimization case study on SImple Reductions. Strategies examined for efficiently mapping reductions onto the ATI Radeon™ HD 5870 GPU and AMD Phenom™ II X4 965 CPU. Taking advantage of properties of the reduction being performed, as well as matching the style of reduction to the hardware platform, can result in performance improvements of up to 15x, compared to naive code.
CMSoft brings an OpenCL accelerated SVM implementation that can be used for general-purpose classification. Support Vector Machine (SVM) is a statistical learning tool considered to be the state-of-the art classifier for many applications today, including medical research and text categorization. Source code is provided showing classification of the MNIST handwritten database.
Phoronix Test Suite Beta 3 adds another handful of Open Computing Language tests for Linux and Mac OS X. Eventually Windows will also be supported. The new test profiles include JuliaGPU, MandelGPU, SmallPT-GPU, and MandelbulbGPU as well an OpenCL test suite has also been added.
The 2010 GPU Technology Conference will run from September 20th - 23rd at the San Jose Convention Center. Pre-conference tutorials will include OpenCL optimization techniques and an introduction to OpenCL. The latest OpenGL features will discussed along with NVIDIA specific extensions. A fifty minute talk on WebGL will introduce this newly emerging standard and help you get started in WebGL development.
Recently, the really interesting discussions from a technological standpoint have been about the use cases for GPU computing. Tom Olson has had some fun recently in his OpenCL blog: where he tries to pick a fight with Jem Davies. Are they both right?
AMD has released an update to their ATI Stream SDK. This brings OpenCL 1.1 support to the table among many other enhancements. Although most of the improvements revolve around OpenCL, there is a host of additional hardware and OS support included also. The ATI Stream SDK v2.2 product is based on a published Khronos Specification, and is expected to pass the Khronos Conformance Testing Process. Current conformance status can be found at www.khronos.org/conformance.