ARM Mali-T622 Overview

Designed for visual computing and using innovative tri-pipe architecture, the ARM® Mali™–T622 GPU delivers significant performance improvements over current Mali graphics processors. The second generation Mali-T600 Graphics Processor Unit (GPU) extends API support to include OpenCL™ Full Profile as well as embedded and Microsoft® DirectX®.

Why Mali-T622?

Mali-T622_block_diagram.jpg

Mali-T622 is the smallest OpenCL 1.1 Full Profile solution, enabling entry level GPU Compute support for the mass market. It brings 50% energy efficiency improvements over first generation Mali-T600 products (Mali-T604). Mali-T622 also supports Renderscript Compute on mass-market phones allowing increased efficiency in processing compute functions such as computational photography and computer vision.

Designed for visual computing and using innovative tri-pipe architecture, the ARM Mali-T622 GPU Compute solution builds upon a track record for high-quality scalable multicore solutions for 2D and 3D graphics. Key APIs supported include OpenGL® ES 1.1, OpenGL ES 2.0 and OpenGL ES 3.0, DirectX® 11 and OpenCL™ 1.1.

The Mali-T622 scales from one to two cores, it is a complete embedded graphics and GPU Compute acceleration platform and enables the development of advanced user interfaces, gaming capabilities and GPU Compute applications for a wide range of tomorrow’s consumer devices, from smartphones to tablets, from DTV to portable gaming and beyond.

Mali-T622 pipelines provide true IEEE double-precision floating-point math in hardware for Full Profile, as well as Embedded Profile, OpenCL support. Production-quality software support, via a single driver stack for all multicore configurations, provides simplified application porting, system integration and maintenance. Multicore scheduling and performance scaling is fully handled within the graphics system, with no special considerations required from the application developer.

The Job Manager within Mali-T600 Series GPUs offloads task management from the CPU to the GPU and enables seamless load balancing across active shader cores. Through ARM coherency and interconnect technology,
compute tasks can conveniently span the CPU and GPU as well as other available compute resources, accessing data more efficiently during shared processing in a heterogeneous system.

Specifications

     

FeaturesValueDescription
Anti-Aliasing

4xFSAA

16xFSAA

4x Full Scene Anti-Aliasing (FSAA) with minimal performance drop
API Support

OpenGL® ES 1.1, 2.0, 3.0

OpenCL™ 1.1

DirectX® 11

Renderscript™

Full support for next-generation and legacy 2D/3D graphics applications
Bus InterfaceAMBA®4 ACE-LITECompatible with a wide range of bus interconnect and peripheral IP
L2 CacheConfigurable 32kB-256kB32KB recommended per core
Memory SystemVirtual MemoryBuilt-in Memory Management Unit (MMU) to support virtual memory
Multi-Core Scaling1 to 2 coresA single IP covering a range of mid to high-end product requirements and price/performance points
Adaptive Scalable Texture CompressionLow dynamic range (LDR) and high dynamic range (HDR).  Supports both 2D and 3D imagesASTC offers a number of advantages over existing texture compression schemes by improving image quality, reducing memory bandwidth and thus energy use.
Anonymous
Graphics & Multimedia blog