Tuesday, May 21, 2013

Vivante demos GPU solutions shipping in 100+ million devices

USA: Vivante Corp. announced its participation in the Multicore Developers Conference in Santa Clara, CA  where it is demonstrating its smaller, faster, cooler GC series IP graphics cores.

Found in many of today's smartphones, tablets, set top boxes and other devices, the GC graphics family is designed from the ground up to accelerate mass market mobile and embedded products, including scalable multi-core applications like automotive, military, medical and others.

Supporting the latest versions of Android, Linux, QNX, and Microsoft Windows, Vivante's product portfolio includes performance leading technologies in 3D, 2D/Composition Processors, GPGPU, and vector graphics. Vivante cores use a unified driver architecture compatible with industry-standard application programming interfaces like OpenGL ES 3.0/2.0, desktop OpenGL, OpenCL, OpenVG, Microsoft DirectX, Google Renderscript Compute, and other standard APIs.

Robust features built into the GPU cores include:

* World's Smallest Licensable GPU Core designed for OpenGL ES 3.0: 3.5 mm2 total silicon footprint (including memories) in TSMC 40nm LP process for area sensitive designs that does not sacrifice on features or performance. Cores can be optionally configured with an integrated Composition Processing Core (CPC) which seamlessly accelerates OS level composition tasks at a fraction of the power with only 10 percent GPU silicon area increase.

* Single/Dual/Quad/Eight GPU Configurations: ScalarMorphic cores scale from highly silicon optimized single GPU cores to performance optimized multi-GPU nodes, all with best-in-class performance and available hardware support for security and OS virtualization.

* Single Software Stack: Single unified software stack supports all Vivante cores across the full range of graphics and compute APIs, features, and performance levels.

* Battery Saving Innovations: Extreme low power microarchitecture and smart grid technologies for leading mW per MHz/GHz and dynamic power control.

* Memory Efficient: Memory bandwidth modulation optimized for contemporary vertex intensive applications and pixel rich displays including support for the latest compression standards in OpenGL ES 3.0 and optionally the recently announced Khronos ASTC – Adaptive Scalable Texture Compression extension.

* Heterogeneous Platform Architecture: GC cores designed for hybrid computing systems like HSA using AMBA ACE-Lite (CPU – GPU cache coherency) and the latest Stream Interface.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.