Submit a News Story

Gpgpu tagged stories

OpenCL Installable Client Driver (ICD) Loader

The OpenCL ICD extension (cl_khr_icd) allows multiple implementations of OpenCL to co-exist on the same system. The OpenCL ICD Loader Library allows applications to choose a platform from the list of installed platforms and dispatches OpenCL API calls to the underlying implementation. Source code for the ICD loader library is available in the Khronos registry. Consult LICENSE.txt in the tarball for full terms and conditions.
Read More

Whitepaper on leveling the playing field for processors with OpenCL

The question of whether OpenCL will eventually help displace GPGPU, by facilitating "GP-something-else" - "general-purpose" accelerators which aren't like GPUs is discussed in a whitepaper by Yosef Kreinin. (via Adapteva)
Read More

Beignet: OpenCL/GPGPU Comes For Ivy Bridge On Linux

Over the past several months there's been a new Intel OpenCL Linux effort: Beignet. The open-source Beignet is an OpenCL/GPGPU implementation targeting Ivy Bridge hardware and newer.
Read More

Intel to give HD4000 a performance boost with OpenCL 1.2

New drivers for Intel's Ivy Bridge-based third-generation Core processors brings support for OpenCL 1.2 general-purpose GPU (GPGPU) language and a 10 percent performance boost. The HD4000 is the GPU found in Intel's Ivy Bridge-based third-generation Core processors.
Read More

The GPGPU Continuum on April 25th 2013 in Haifa

The GPGPU Continuum from mWatts to peta flops workshop addresses the current and future challenges of the GPGPU community. This workshop aims to address two important directions of future GPGPU based systems: at the high-end, it will focus on using GPUs as part of clusters, clouds, compute farms, etc., that aim to achieve peta flops of computations. At the low-end, it will focus on using GPUs as part of mobile devices, which limits the power consumption of the GPU to mWatts. We conclude the workshop with a panel discussion on the differences and similarities and on the challenges each domain faces.
Read More

Harlan targets complexity for GPGPU programming and compiles to OpenCL

HPC programmers who are tired of managing low-level details when using OpenCL or CUDA to write general purpose applications for GPUs (GPGPU) may be interested in Harlan, a new declarative programming language designed to mask the complexity and eliminate errors common in GPGPU application development. The idea with Harlan is to keep developers focused on the high-level HPC programming challenge at hand, instead of getting bogged down with the nitty gritty details of GPU development and optimization. Harlan's syntax is based on the language Scheme, and compiles to Khronos Group's OpenCL.
Read More

Qualcomm announces commercial 64-bit Octa-Core chipset with OpenCL and OpenGL ES 3 support

Qualcomm has added the Snapdragon 610 and 615 chipsets to the Snapdragon 600 tier for high-end mobile computing devices. Both new chipsets integrate Qualcomm Technologies’ 3rd Generation LTE modem, supporting Category 4 data rates for new requirements such as LTE-Broadcast and LTE Dual SIM Dual Active (DSDA). The chipsets also feature Qualcomm Technologies’ Adreno 405 GPU which support the latest mobile graphics APIs like DirectX 11.2 and OpenGL ES 3.0 with added support of hardware accelerated geometry shading and hardware tessellationfor more detailed, realistic mobile games and visually stunning user interfaces. The Adreno 405 also supports Full Profile OpenCL for superior GPGPU compute, video and image processing.
Read More

FinalWire Introduces AIDA64 v4.50 with new OpenCL benchmarks

FinalWire has announced AIDA64 v4.50. The new AIDA64 release introduces a multi-threaded OpenCL GPGPU AES-256 benchmark to measure cryptographic processing performance of modern graphics processors and APUs.
Read More

Boost.Compute v0.3 proposed header-only C++ library based on OpenCL

Boost.Compute is a header-only C++ library for GPGPU and parallel-computing based on OpenCL. This proposed library for Boost is available on GitHub and instructions for getting started can be found in the documentation. Are you interested in Parallel Computing? The Boost.Compute project is currently looking for additional developers. Boost.Compute is not yet an official Boost library.
Read More

Page 8 of 8 pages ‹ First  < 6 7 8

safety