by Anton Shilov
08/02/2012 | 10:25 PM
ARM has submitted its ARM Mali-T604 graphics processing unit (GPU) for OpenCL 1.1 full profile conformance with Khronos, a standard-setting organization. The conformance will ensure that system-on-chips with ARM Mali-T604 graphics technology will be able to do general-purpose computing using GPU computing units.
The Mali-T600-series of GPUs have been designed from the outset for GPU computing, including full support for work-groups and synchronization barriers. Building on a scalable multicore, multi-pipeline architecture design, the Mali-T600 GPU includes a number of advanced features. In particular, native scalar and vector operations for OpenCL's integer and floating point data types (including 64-bit); support for static and dynamic compilation; hardware accelerated image and sampler data types; fast atomic operations and compliance to IEEE754-2008 precision requirements.
Building on the heritage of Mali-400 MP, the Mali-T604 is ARM’s fourth-generation GPU and is the first member of a new family of GPUs based on the Midgard architecture which all use a common software driver. The Mali-T604 delivers up to 5x performance improvement over previous Mali graphics processors and is specifically designed to be compliant with GPGPU-based applications that use OpenCL. ARM's Mali-T604 can feature up to four stream processors and also features new techniques that reduce memory bandwidth consumption, by up to 30%.
GPU computing is a trend where the computational performance of the GPU, historically used for graphics, is harnessed to augment the main processor (or CPU) for certain applications where the GPU architecture will be more effective. The Open Computing Language (OpenCL) is an important open standard programming framework for portable, parallel computation that enables the CPU and GPU in a system to work faster and more efficiently together.
ARM is the first GPU intellectual property (IP) supplier to submit conformance for full profile OpenCL, bringing desktop GPU computing features to the mobile, embedded and smart-TV markets. This will help bring visual computing to life by providing developers with a consistent and effective platform on which to build their designs, and tangible benefits to the end user in terms of improved performance and battery life of their devices.